(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Transactions of the Royal Society of Edinburgh"

TRANSACTIONS 



OF THE 



EOYAL SOCIETY OF EDINBURGH. 



S.V-C.V^ 



TRANSACTIONS 



OF THE 



ROYAL SOCIETY 



OF 



EDINBURGH. 



VOL. XLVIII. 



EDINBURGH : 

PUBLISHED BY ROBERT GRANT & SON, 107 PRINCES STREET, 
AND WILLIAMS & NORGATE, 14 HENRIETTA STREET, COVENT GARDEN, LONDON. 



MDCCCCXIII, 




No. 



I. 


Published 


May 16, 1911. 


No. 


XVIII. Published 


II. 


)> 


May 16, 1911. 


» 


XIX. 


III. 


n 


July 3, 1911. 


55 


XX. 


IV. 


>> 


August 9, 1911. 


55 


XXI. 


V. 


5 J 


August 30, 1911. 


)3 


XXII. 


VI. 


;> 


October 20, 1911. 


55 


XXIII. 


VII. 


j) 


November 7, 1911. 


J) 


XXIV. 


VIII. 


>> 


December 18, 1911. 


>' 


XXV 


IX. 


,, 


January 19, 1912. 


55 


XXVI. 


X. 


)) 


February 28, 1912. 


5) 


XXVII. 


XI. 


>? 


April 1, 1912. 


)» 


XXVIII. 


XII. 


>j 


April 4, 1912. 


5) 


XXIX. 


XIII. 


51 


April 13, 1912. 


55 


XXX. 


XIV. 


)) 


May 28, 1912. 


,, 


XXXI. 


XV. 


J) 


May 23, 1912. 


>> 


XXXII. 


XVI. 


,, 


July 3, 1912. 


>; 


XXXIII. 


XVII. 


1' 


July 18, 1912. 


)> 


XXXIV. 



August 26, 1912. 
August 28, 1912. 
August 17, 1912. 
August 30, 1912. 
September 6, 1912. 
September 21, 1912. 
November 15, 1912. 
November 18, 1912. 
December 14, 1912. 
January 9, 1913. 
January 10, 1913. 
December 23, 1912. 
February 8, 1913. 
February 17, 1913. 
February 19, 1913. 
March 24, 1913. 
April 3, 1913. 



CONTENTS. 



PART I. (1911-12.) 

NUMBER PAGE 

I. The Nemertines of Millport and its Vicinity. By J. Stephenson, 
M.B., D.Sc. (Lond.); Major, Indian Medical Service; Professor of 
Biology in the Government College, Lahore. (With One Plate), . 1 

II. On some littoral Oligochceta of the Clyde. By J. Stephenson, M.B., 
D.Sc. (Lond.); Major, I. M.S. ; Professor of Biology in the Govern- 
ment College, Lahore. (With Two Plates), . . . .31 

III. Les Mousses de V Expedition nationale antarctiaue ecossaise. Par 

Jules Caroot. (Avec trois Planches), . . . .67 

IV. The Pharmacological Action of Harmine. By James A. Gtjnn, 

M.A., M.D., D.Sc. (From the Pharmacology Laboratory of the 
University of Edinburgh), . . . . .83 

V. On the Resistance to Flow of Water through Pipes or Passages having 
Divergent Boundaries. By Professor A. H. Gibson, D.Sc, University 
College, Dundee, . . . . . . .97 

VI. The Significance of Maximum Specific Electrical Conductivity in 

Chemistry. By Professor John Gibson, Ph.D., . . .117 

VII. Nuclear Osmosis as a Factor in Mitosis. By A. Anstruther Lawson, 
Ph.D., D.Sc, F.L.S., Lecturer in Botany, University of Glasgow. 
(With Four Plates), . . . . . . .137 

VIII. On the Structure and Affinities of Metaclepsydropsis duplex 
(Williamson). By W. T. Gordon, M.A., B.A., D.Sc, Falconer 
Fellow of Edinburgh University, Lecturer in Palaeontology, 
Edinburgh University. (With Four Plates), . . .163 

IX. Scottish National Antarctic Expedition : Observations on the Anatomy 
of the Weddell Seal (Leptonychotes Weddelli). Part II. By David 
Hepburn, M.D., CM., Professor of Anatomy, University College, 
Cardiff (University of Wales), . . . . .191 



vi CONTENTS. 

NDMBHB PAGE 

X. The Influence of the Ratio of Width to Thickness upon the Apparent 
Strength and Ductility of Flat Test-bars of Mild Steel. By 
\Y. Gordon, B.Sc, A.M.I. Mech.E., Lecturer in Mechanical Engineer- 
ing in Leith Technical College; and G. H. Gulliver, B.Sc, 
A.M. I. Mech.E., Lecturer in Engineering in the University of 
Edinburgh, . . . . . . . .195 



-&- 



XI. A Monograph on the general Morphology of the Myxinoid Fishes, 
based on a study of Myxine. Part IV. — On some Peculiarities 
of the Afferent and Efferent Branchial Arteries of Myxine. By 
F. J. Cole, D.Sc., Oxon., Professor of Zoology, University College, 
Reading. (With One Plate), . . . . .215 



PART II. (1911-12.) 

XII. TJie Effect of changing the Daily Routine on the Diurnal Rhythm in 
Body Temperature. By Sutherland Simpson, M.D., D.Sc. (From 
the Physiological Laboratory, Medical College, Cornell University, 
Ithaca, N.Y., U.S.A.) (With Thirteen Figures in the Text), . 231 

XIII. On the Carboniferous Flora of Berwickshire. Part 1. — Stenomyelon 

Tuedianum Kidston. By R. Kidston, LL.D., F.R.S. ; and D. T. 
Gwynne-Vaughan, M.A., Professor of Botany, Queen's University, 
Belfast. (Plates I. -IV.), . . . . . .263 

XIV. The Cephalopoda of the Scottish National Antarctic Expedition. By 

William Evans Hoyle, M.A., D.Sc, . . . . 273 

X V. On Branchiura sowerbyi Beddard, and on a new species of Limno- 
drilus with distinctive cliaracters. By J. Stephenson, M.B., D.Sc. 
(Lond.); Major, Indian Medical Service; Professor of Biology, 
Government College, Lahore. (With Two Plates), . . .285 

XVI. The Tunicata of the Scottish National Antarctic Expedition, 1902- 
1904. By W. A. Herdman, D.Sc, F.R.S., Professor of Zoology in 
the University of Liverpool. (With One Plate), . . . 305 

XVII. Scottish National Antarctic Expedition : Observations on the Anatomy 
of the Weddell Seal (Leptonychotes Weddelli). Part III. By David 
Bepburn, M.D., CM., Professor of Anatomy, University College, 
Cardiff (University of Wales), . . . . .321 



CONTENTS. vii 

NUMBER PAGE 

XVIII. The Marine Mollusca of the Scottish National Antarctic Expedition. 
Part II. By James Cosmo Melvill, M.A., D.Sc, F.L.S. ; and 
Robert Standen, Assistant Keeper, Manchester Museum. (With 
One Plate), . . . . . . . .333 

XIX. The Brachiopoda of the Scottish National Antarctic Expedition (1902 
to 1904). By J. Wilfrid Jackson, F.G.S., Assistant Keeper, 
Manchester Museum. (With Two Plates), . . . .367 

XX. The Equilibrium of the Circular-Arc Bow-Girder. By Professor 
A. H. Gibson, D.Sc, A.M.I.C.E., University College, Dundee. (With 
Eleven Diagrams), . . . . . .391 

XXI. Experiments to show how Failure under Stress occurs in Timber, its 
Cause, and Comparative Values of the Maximum Stresses induced 
when Timber is fractured in Various Ways. By Angus R. Fulton, 
B.Sc, A.M.Inst.C.E., Engineering Department, University College, 
Dundee. (With Eight Plates and Five Text Illustrations), . .417 

XXII. The Cestoda of the Scottish National Antarctic Expedition. By 
John Rennie, D.Sc. ; and Alexander Reid, M.A., University of 
Aberdeen. (With Two Plates), . . . . .441 

XXIII. The Amphipoda of the Scottish National Antarctic Expedition. By 
Chas. Chilton, M.A., D.Sc. (N.Z.), M.B., CM. (Edin.), Hon. LL.D. 
(Aber.), F.L.S. ; Professor of Biology, Canterbury College, New 
Zealand. (With Two Plates), . . . . .455 



PART III. (1912-13.) 

XXIV. The Entomostraca of the Scottish National Antarctic Expedition, 
1902-1904. By Thomas Scott, LL.D., F.L.S. (With Fourteen 
Plates), ........ 521 

XXV. A Study in Chromosome Reduction. By A. Anstruther Lawson, 
Ph.D., D.Sc, F.L.S. ; Lecturer in Botany, University of Glasgow. 
(With Three Plates), . . . . .601 

XXVI. Temperature Observations in Loch Earn. With a further Contribution 
to the Hydrodynamical Theory of the Temperature Seiche. By 
E. M. Wedderburn, W.S. . . . . . .629 



VII 1 



dONTRNTN. 



XW'II. Multiple Neuromata of the Central Nervous System : their Structure 
and Histogenesis. By the late Alexander Bruce, M.D., LL.D., 
F.R.C.P.E. ; and James W. Dawson, M.D. (Carnegie Research 
Fellow). (With Eight Plates), . . . . -697 



PART IV. (1912-13.) 

XXVIII. The Loss of Energy at Oblique Impact of Two Confined Streams 
of Water. By Professor A. H. Gibson, D.Sc., University College, 
Dundee. (With Five Diagrams), ..... 

XXIX. On Rhetinangium arberi, a new genus of Cycaclofilices from the 
Calciferous Sandstone Series. By W. T. Gordon, M.A., B.A., D.Sc, 
Lecturer in Palaeontology, Edinburgh University. (With Three 
Plates), ........ 

XXX. Scottish National Antarctic Expedition: Observations on the 
Anatomy of the Weddell Seal (Leptonychotes Weddelli). Part IV. : 
The Brain. By David Hepburn, M.D., CM., Professor of Anatomy, 
University College, Cardiff (University of Wales). (With One 
Plate), ........ 

XXXI. Scottish National Antarctic Expedition: A Contribution to the 
Histology of the Central Nervous System of the Weddell Seal 
{Leptonychotes weddellii). By Harold Axel Haig, M.B., B.S. 
(Lond.), M.R.C.S. (Eng.), L.RC.P. (Lond.), Lecturer in Histology 
and Embryology, University College, Cardiff. (With Two Plates 
and Nine Text-Figs.), ...... 

XXXII. Jurassic Plants from Cromarty and Sutherland, Scotland. By 
A. C. Seward, M. A., F.R.S., Professor of Botany, Cambridge; and 
N. Bancroft, B.Sc, F.L.S., Newnham College, Cambridge. (Plates I. 
and II. ; Text-Figs. 1-6), ...... 

XXXIII. The Right Whale of the North Atlantic, Balgena biscayensis : its 
Skeleton described and compared with that of the Greenland Right 
Whale, Balsena mysticetus. By Principal Sir Wm. Turner, K.C.B., 
D.C.L., F.R.S., President of the Society, Knight of the Royal Prussian 
Order Pour le Merite. (With Three Plates, and Figures in Text) 

XXXIV. The Geology of South- Eastern Kincardineshire. By Robert 
Campbell, M. A., D.Sc, Lecturer in Petrology in the University of 
Mii,bui-li. (Will. Tli i-ec Plat<||), . 

Index, . • • 

t m I) 



799 



813 



827 



849 



867 



889 

923 
961 






^-.TRANSACTIONS 



OF THE 



ROYAL SOCIETY OF EDINBURGH. 

VOLUME XLVIII. PART I.— SESSION 1911-12. 



S 



CONTENTS. 

PAGE. 

I. The Nemertines of Millport and its Vicinity. By J. Stephenson, M.B., D.Sc. (Lond.); 
Major, Indian Medical Service ; Professor of Biology in the Government College, Lahore. 
Communicated by Professor DArcy W. Thompson, C.B. (With One Plate), . . 1 

(Issued May 16, 1911.) 

II. On some littoral Oligocheela of the Clyde. By J. Stephenson, M.B., D.Sc. (Lond.) ; Major, 
I.M.S. ; Professor of Biology in the Government College, Lahore. Communicated by 
Professor D'Arcy W. Thompson, C.B. (With Two Plates), .... 31 

(Issued May 16, 1911.) 

III. Les Mousses de V Expedition nationale antarctique ecossaise. Par Jules Cardot. Presents par 

le Professeur I. Bayley Balfour, M.D., F. U.S. (Avec trois Planches), . . .67 

(Issued July 3, 1911.) 

IV. The Pharmacological Action of Harmine. By James A. Gunn, M.A., M.D., D.Sc. (From the 

Pharmacology Laboratory of the University of Edinburgh), . . . .83 

(Issued August 9, 1911.) 

V. On the Resistance to Flow of Water through Pipes or Passages having Divergent Boundaries. 
By Professor A. H. Gibson, D.Sc, University College, Dundee. Communicated by Professor 
W. Peddie, D.Sc, .......... 97 

(Issued August 30, 1911.) 

VI. The Significance of Maximum Specific Electrical Conductivity in Chemistry. By Professor 

John Gibson, Ph.D., . . . . . . . . .117 

(Issued October 20, 1911.) 

VII. Nuclear Osmosis as a Factor in Mitosis. By A. Anstruther Lawson, Ph.D., D.Sc, F.L.S., 

Lecturer in Botany, University of Glasgow. (With Four Plates), . . . .137 

(Issued November 7, 1911.) 

VIII. On the Structure and Affinities of Metaclepsydropsis duplex ( Williamson). By W. T. Gordon, 
M.A., B.A., D.Sc, Falconer Fellow of Edinburgh University, Lecturer in Palaeontology, 
Edinburgh University. Communicated by Professor James Geikie, D.C.L., LL.D., etc 
(With Four Plates), . . . . . . . . .163 

(Issued December 18, 1911.) 

IX. Scottish National Antarctic Expedition : Observations on the Anatomy of the Weddell Seal 

(Leptonychotes Weddelli). Part II. By David Hepburn, M.D., CM., Professor of 
Anatomy, University College, Cardiff (University of Wales), . . . .191 

(Issued January 19, 1912.) 

X. The Influence of the Ratio of Width to Thickness upon the Apparent Strength and Ductility of 
Flat Test-bars of Mild Steel. By W. Gordon, B.Sc, A.M.I.Mech.E., Lecturer in Mechanical 
Engineering in Leith Technical College, and G. H. Gulliver, B.Sc, A.M.I.Mech.E., 
Lecturer in Engineering in the University of Edinburgh, . . . . .195 

(Issued February 28, 1912.) 

XI. A Monograph on the general Morphology of the Myxinoid Fishes, based on a study of Myxine. 
Part IV. — On some Peculiarities of the Afferent and Efferent Branchial Arteries of Myxine. 
By F. J. Cole, D.Sc. Oxon., Professor of Zoology, University College, Reading. Com- 
municated by Dr E. H. Traqu air, F.R.S. (With One Plate), . . . .215 

(Issued April 1, 1912.) 



EDINBURGH: 

PUBLISHED BY ROBERT GRANT & SON, 107 PRINCES STREET, 

AND WILLIAMS & NORGATE, 14 HENRIETTA STREET. COVENT GARDEN, LONDON. 



MDCCCCXII. 
Price Twenty-two Shillings and Ninepence. 




TRANSACTIONS. 



I. — The Nemertines of Millport and its Vicinity. By J. Stephenson, M.B., D.Sc. 
(Loncl.) ; Major, Indian Medical Service ; Professor of Biology in the Government 
College, Lahore. Communicated by Professor D'Arcy W. Thompson, C.B. (With 
One Plate.) 

(MS. received July 4, 1910. Read November 21, 1910. Issued separately May 16, 1911.) 

The Nemertines of Millport and its Vicinity. 

The present communication deals with the Nemertines of Millport and its vicinity, 
the study of which occupied part of the time I spent at the Millport Biological Station 
during May and June 1909. The material was not completely worked out during my 
stay there, and most of the section -cutting was left till after my return to India. I 
wish to acknowledge here the uniform courtesy and kindness which I received from 
Mr R. Elmhirst, the Superintendent of the Station, to whom my thanks are due for 
the readiness with which he placed the resources of the institution at my disposal, and 
for help in other ways. 

The number of species examined was twelve ; this list is probably fairly complete 
as regards the Millport littoral forms, though I have little doubt that other species will 
be found to occur at greater depths. 

The results which I believe have been attained may be summarised as follows : — 

(1) Of the forms examined, two — an Amphiporus and a Micrura — appear to be new; 
their descriptions are given at length. 

(2) Not less interesting has been the working over of certain species already fairly 
well known. I have not, of course, given full descriptions of these, but have contented 
myself with drawing attention to variations, which are sometimes marked, between my 
observations and the accounts of previous writers ; and in a few cases I have been able 
to add facts which have escaped previous record. 

(3) It is in a consideration of the variations just alluded to that the chief value of 
the work may be found to lie. The great variability of the class is evidenced in the 
first place by my own observations ; many of the forms examined are very common, 
and a large number of specimens passed under my investigation. In the second place, 

TRANS. ROY. SOC. EDIN., VOL. XLVIII, PART I. (NO. 1). 1 



2 DR J. STEPHENSON ON 

and even more strikingly, it is brought out by comparisons with the descriptions of 
those who have worked in other localities. 

The authors whose works 1 have used for the purpose of comparison are MacIntosh 
(10), Joubin (7, 8), and Burger (3, 5). Macintosh's monograph (1873), with its 
valuable coloured plates, represents the chief attempt at an exhaustive account of the 
Nemertine fauna of Great Britain; Joubin's works (1890-1894), though without the 
same wealth of illustration, represent a similar undertaking for the coasts of France ; 
Burger's great monograph (1895) describes in detail the rich Nemertine fauna of Naples, 
with the aid of a beautiful series of plates, and more recently (1904) the same author 
has given in the Tierreich a general systematic account of the whole class. 

It would, of course, have been better if a larger number of authors could have been 
used for purposes of comparison, and especially if the original descriptions of the several 
species could have been utilised for this purpose ; but my recent stay in England, 
and, with it, my access to zoological literature, was unfortunately of short duration. I 
believe, nevertheless, that, though thus incomplete, the comparisons I have been able to 
make between the various species, as met with by myself and by previous writers 
respectively, are not without value and interest. 

The variability of Nemertines as a class has long been known as regards colour, 
and in specific diagnoses the colour, though recognised as one of the principal points 
requiring description, is usually stated in wide terms. 

It seems to me, however, that not only colour, but also other characters commonly 
used for purposes of specific distinction, are variable in a high degree. Thus this appears 
to be the case with regard to (a) the shape of the head (compare below, under Lineus 
longissimus, L. ruber, Prostoma candidum) ; (b) the degree in which the head is marked 
off from the body (Lineus longissimus) ; (c) the length of the specimens (the Millport 
examples of Prostoma candidum and Emplectonema neesii as compared with Macintosh's 
specimens from the East Coast) ; (d) the number and arrangement of the eyes (see the 
comparison of the several authors' descriptions of Lineus longissimus, Amphiporus 
pulcher, and Emplectonema gracile given below, and compare the Millport specimens 
of Amphiporus lactijloreus and also, as regards size of eyes, Prostoma candidum with 
Burger's descriptions) ; (e) the cephalic grooves (compare the descriptions of Amphi- 
porus lactijloreus and A. pulcher here given with those of Joubin) ; if) the arrangement 
of the musculature (Millport and Naples specimens of Tubulanus annulatus). 

Two variations remain, which seem to demand more than mere mention. The first 
of these concerns (g) the shape of the basis of the stylet. This feature one would 
naturally be inclined to look on as a definite morphological character, capable, if any- 
thing were so, of affording a criterion of specific distinction. Thus Joubin (9) writes 
concerning it : "II peut-etre plus ou moins long, etroit ou renfle, pourvu d'ailerons 
lateraux ou presque carre ; tous les caracteres sont utilises dans la determination des 
especes." The author who most often describes this structure in species possessing it is 
Burger, and it is somewhat surprising to find that in most cases (Amphiporus lacti- 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 3 

jloreus, A. pulcher, Prostoma candidum, Oerstedia dorsalis) the Millport specimens 
show a wide divergence from the Naples forms. It is curious, too, that one particular 
difference repeats itself in all these cases — the Naples forms show a marked constric- 
tion about the middle of the basis, which is scarcely to be observed at all in the same 
species as found at Millport. In the case of Amphiporus lactifioreus the shape varies 
even among the Millport specimens. 

Attention may here be called to the fact that Oxner (12) has also found the shape 
of the basis to be variable in Prosorochmus delagei, a new species described by him. 
He says : " La forme et les dimensions du socle peuvent subir de nombreuses variations. 
Cette variation est un trait caracteristique pour P. delagei." The figure shows that 
the basis may either present a constriction or not, that its posterior end may be either 
wider or narrower than the anterior end, and that its size may vary, both relatively to 
the size of the stylet, and also absolutely. He, however, supposes, as we have seen, that 
this variability is a special feature of this particular form. 

The fact that Burger in different works gives two different shapes for the basis of 
the stylet in Prostoma flavidum is presumably also to be explained by the variability 
of this structure. In the Naples monograph, the figure shows it as somewhat dumbbell- 
shaped — that is, with rounded ends of equal size, and a constriction in the middle ; 
in the Tierreich, however, it is described as "Der kegelformige, kaum in der mitte 
eingeschniirte Sockel." 

Lastly, (h) the number of proboscis nerves furnishes an example of another definite 
morphological character which might have been supposed to be constant and capable 
of furnishing a criterion of specific distinction. Yet Amphiporus pulcher is stated by 
Burger in the Tierreich to possess ten proboscis nerves, while the Millport specimens 
examined for this character showed twelve. 

The importance of a recognition of the variability of the above-mentioned characters 
lies in the fact that they are among those which are most frequently used for purposes 
of diagnosis. With regard to the shape of the head, for example, the details of which 
are used in many genera as specific distinctions, and in some cases take the chief, or at 
least a leading part, among the criteria of diagnosis, one constantly meets, combined in 
various ways, such expressions as spoon-shaped, lancet-shaped, egg-shaped ; marked off, or 
not marked off from the body; wider than, slightly wider than, or not wider than the body. 

Consider for a moment such specific diagnoses as these [Tierreich) — (i.) Prostoma 
vermicular is (Quatr.) : " Vorderende verbreitert, Hinterende etwas verjtingt. Kopf 
spatelformig, nicht vom Rumpf abgesetzt. Mit 4 kleinen Augen, die im Rechteck 
stehen. L. 12-15, Br. kaum 1 mm." # 

(ii.) Prostoma flavidum (Ehrbg.) : "Kopf vorn abgerundet, spatelformig, nicht 
vom Rumpf abgesetzt. Rosenfarben, Seitenrander durchscheinend. Mit 4 einfachen, 
sehr kleinen Augen, die im Rechteck stehen. Angriffsstilett ein wenig langer als der 

* P. vermicularis is divided, certainly, into three sub-species, which are distinguished among themselves by various 
markings or the absence of them. The above, however, constitutes the whole of the specific diagnosis. 



4 DR J. STEPHENSON ON 

kegelformige, kaum in der mitte eingeschniirte Sockel ; beide schlank. Reservestilet- 
taschen mit je 3 Reservestiletten. L. etwa 13, Br. 0*75 mm." 

Seeing that no particulars as to the stylet or its basis are given in the case of 
P. vermicularis, when allowance is made for slight variation of the shape of the head, 
length, etc., does any tangible difference remain ? 

Or, again — (i.) Cephalothrix rufifrons (Johnst.) : " Weisslich. Mit 2 kleinen rot- 
blauen Pigmentflecken an der kopfspitze. L. 30-40, Br. 0*5 mm." (ii.) Cephalothrix 
bipunctata (Biirg.) : " Ockergelb, Kopf heller. Dicht vor dem Gehirn 2 kleine schwarze, 
seitliche Pigmentflecke. L. 60-100, Br. 1 mm." 

I have considered a similar case at some length below ; but again, in such cases as 
this, when necessary allowances are made, does anything remain ? 

It will, I think, be agreed that the present observations tend to show the inadequacy 
of many of the specific characters in use, and the need for new ones. It seems 
probable that such will have to be sought for in the details of internal anatomy, rather 
than, as has been the case with most of the old ones, anions; the external features. 
That this will make the description and identification of Nemertine species a more 
difficult matter than has hitherto been the case cannot be doubted. One has only to 
compare, for example, the laboriousness of the task of description of Enchytrseid or 
Tubificid species, where every identification requires a complete series of sections. 

(4) The fourth result which I believe has been reached is the union of two species 
of Cephalothrix into one, as just mentioned. I have considered this particular case in 
some detail, since I believe that it is typical of what will have to be done in the future 
in a number of similar cases. I have also, in another place, given reasons which seem 
to me to go far towards establishing the necessity for a similar treatment of two species 
of Prostoma. My observations also, as stated below, support the unification by Joubin 
and by Burger of the two species Linens gesserensis and L. sanguineus, considered by 
MacIntosh as distinct. It seems to me to be beyond doubt that further work on similar 
lines would have similar results in other species also. 

(5) Finally, I may mention here the observations on the physiology of the circu- 
lation in Cephalothrix linearis. In view of the paucity of such physiological observa- 
tions in this class they appear to be of interest. 

The nomenclature adopted throughout is that of the Tierreich. 

Tubulanus annulatus (Carinella annulata) (Montagu). 

Specimens are not common at Millport ; they were found under stones near low- 
water mark at Balloch, and were dredged in 15 fathoms off Ascog Bank. 

Length 3^ 7 inches (85-170 mm.); breadth 1 mm., broadest at the anterior end, 
tapering gradually to a fine point posteriorly. These figures for the length of this animal 
are practically those given by Burger in the Tierreich ; they are larger than those for 
specimens found at Naples (3) (8-10 cm.), though much less than those given by 



THE NEMERTTNES OF MILLPORT AND ITS VICINITY. 5 

Macintosh (7-30 inches) and especially Joubin (7). The latter speaks of specimens 
measuring 80 cm., and sometimes 1*5 metres, and quotes Quatrefages as giving 2 metres. 

The striking coloration has been described in detail by previous writers. Briefly, 
the general colour of the body varies from a rich red to a vandyke brown ; there is a 
white mid-dorsal stripe, a pair of similar lateral stripes, and a number of white trans- 
verse stripes which encircle the whole body ; the mid-dorsal stripe does not reach the 
anterior end of the body, but leaves a red " frontal patch" (Stirnfeld), undivided at the 
anterior tip. 

Points which have either not received mention by previous observers, or in which 
my specimens, which were generally of a brilliant brown colour, differ from theirs, are 
as follow : — 

(i.) The frontal patch was always less brilliant, or lighter in tint, than the rest of 
the coloured surface of the dorsum. This was evidently not the case in Burger's 
specimens ; the frontal patch is stated to have the same appearance as the general 
coloured surface of the body, and it is so shown in the plate. 

(ii.) The ventral surface was of a lighter brown than the dorsal ; it was sometimes 
white in the greater part of its extent, becoming light brown posteriorly, though here 
still much paler than the dorsum. 

(iii.) The mid-dorsal and lateral white stripes consisted of a dull white ground with 
a longitudinal stripe of a more intense opaque white down the middle of each ; the 
transverse stripes also had mostly a similar appearance. 

(iv.) The genital apertures appeared as a series of whitish spots, dorso-laterally 
placed, beginning some distance behind the head, at first in a single row on each 
side, but more numerous posteriorly. 

The cephalic grooves slightly notch the lateral borders of the head a little in front 
of the level of the mouth : thence followed inwards on the dorsal surface they have a 
slightly sinuous course, at first convex forwards, then concave, ending near the median 
line ; on the ventral surface they are continued directly inwards from the lateral notch, 
almost meeting each other in the middle line. 

The cerebral organs are present in this genus as mere grooves, not as canals or 
sacs. MacIntosh, however, in the general account of the cephalic sacs of the Anopla 
(10), says : "Just in front of the external border of the curved dorsal groove on the 
snout of Carinella annulata is an ovoid body apparently homologous with the fore- 
going, but I have not yet been able to trace its anatomy, on account of the opacity 
of the cutaneous tissues in this animal." In his description of this species he says : 
" There is a curved streak in the bend of each ciliated furrow on the dorsum, perhaps 
in connection with the cephalic sac." The corresponding figure (10, pi. xvii., fig. 24) 
shows an ovoid structure of fair size in the bend of the groove on the dorsal surface, 
near the lateral margin, a little in front of the level of the mouth. Burger states (5) 
that in this species, " die cerebral Organe sind kugelige Gebilde." 

The true state of affairs appears to be as follows : On the dorsal surface of the 



6 DR J. STEPHENSON ON 

head, just in front of the cephalic grooves and in their concavity, there is to be seen 
a circular area, rather indefinite in its limits, which is of a lighter brown than the 
surrounding parts ; with an ordinary microscope nothing further is to be made out, 
but with a binocular it appears that this area is slightly depressed below the general 
surface (figs. 1,2). It is apparently this area which corresponds to the oval outline 
in Macintosh's figure. 

Sections show that there is no cerebral organ of the usual type — that is, that there 
are no canals or sacs ; the depressed areas just mentioned are, however, recognisable 
as areas of the epidermis with cells and ciliation like those of the cephalic grooves ; 
the areas are sunk below the general surface, and, owing probably to contraction in 
killing and fixing, their margins are much better defined than in life and project 
inwards towards the centre of the area (fig. 3). 

The species, therefore, does not form an exception — at least not such a marked 
exception as would be inferred from previous descriptions — to the general rule for the 
genus in regard to the cerebral organs. 

A well-developed inner (splanchnic) circular muscular layer is characteristic of 
the family Tubulanidse (Carinellidse) ; but this particular species is stated to form 
an exception to the rule. Thus Burger (5) writes : " Die innere Ringmuskelschicht 
ist ziemlich schwach entwickelt, nur das dorsale muskelfaserkreuz ist vorhanden." 
My specimens must differ markedly from those on which this statement is founded ; 
they agree with others of the family in having a well-developed inner circular 
layer, while, on the other hand, the dorsal crossing of the fibres (as represented, e.g., 
for Tuhulanus superbus, 5, p. 3, fig. 1 ; 1, p. 169, fig. viii.) is not recognisable. 

A feature, the description of which I have not met with elsewhere, is the character of 
the epithelial lining in the most anterior part of the proboscis. Almost immediately 
behind the spot where the proboscis becomes free within its sheath the lining epithe- 
lium of its dorsal, and somewhat less markedly of its ventral, wall is formed by a compact 
mass of tall and very narrow cells. This patch, elongated transversely to a semicircle 
which comprises the dorsal half of the circumference of the proboscis, is narrow antero- 
posteriorly, so that in a longitudinal section it appears as a cushion-like or fan-like 
projection (figs. 4, 5). The cells of which it is composed stain deeply, and are of a very 
different appearance from the much lighter, more loosely arranged, and more ragged- 
looking cells which succeed them ; two layers of nuclei are visible, one near the base, 
the other at about the middle of the ' cushion.' The appearances are similar, but the 
cells are not so high, on the ventral wall of the tube (fig. 4). 

Cep/ialothrix linearis (J. Rathke). 

Under the name Cephalothrix linearis, MacIntosh has united certain forms which 
other writers have separated under the names C. linearis and C. rufifrons (or bioculata). 
It is a question how far the Millport specimens bear out this separation. 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 7 

Macintosh's description (10) of C. linearis gives the length as three to four inches. 
The colour is stated to be variable ; the animals are said to be in general of a pale 
cream tint, which is sometimes diversified by a yellowish patch on the snout and a 
yellowish tinge in the oesophageal region ; or the pigment — yellowish, orange, or 
reddish — of the snout may be increased towards its tip, the oesophageal region being 
also in such cases of a reddish orange colour. There are said to be no eyes ; and the 
red pigment at the anterior end is stated to be of no specific value. In the specimen 
represented in this author's pi. iv. fig. 5 (the posterior part of the specimen being green), 
the reddish pigment of the snout is shown as extending farther back laterally than in 
the middle line. Nothing is said in the text as to the aggregation of the pigment in 
granules. 

Joubin (8) describes separately the forms with red anterior tip as C. bioculata. He 
states that in these the head is red, the red colour increasing in depth towards the tip 
of the snout ; the pigment occurs in the form of minute granules ; the two eyes, con- 
stituted perhaps by the agglomeration of minute ocelli, are of considerable size, on the 
extreme margin, and placed among the red pigment grains. The same author (7) 
speaks of the presence of two " oculiform points," which may or may not be capable of 
resolution into a number of small eyes, as the feature which distinguishes 0. bioculata 
from C. linearis ; C. bioculata is also said to be shorter than C. linearis. 

Burger (3), also describing C. linearis and C. bioculata separately, states that the 
single specimen of C. linearis met with at Naples had no eyes, nor pigment of any kind 
in the head. C. bioculata is said to be three to four cm. long, and to be colourless or 
whitish with the exception of two very small bright red spots at the end of the head ; 
with weak magnification, there is visible at the anterior border of the head a small black 
pigment spot, which, however, may not be an eye ; and immediately behind this are two 
larger, sharply defined, round reddish spots in which coerulean-blue pigment is inter- 
spersed. It is difficult to see much of this in the corresponding figure, which, being on 
a small scale, does not even show the separation of the red areas of the two sides. 
Burger compares his specimens of C. bioculata with those of Macintosh's specimens 
(of C. linearis) which possessed the red pigment ; he considers the two to be the same 
form, though he recognises that Macintosh's specimens showed a more diffuse dis- 
tribution of the pigment than those found by him at Naples. 

In the Tierreich (5), Burger defines C. linearis as being white, often with a yellow 
tinge, as having no pigment spots, and as being 100-150 mm. in length. C. rujifrons 
( = bioculata), on the other hand, is whitish, with two small red-blue pigment spots at 
the end of the head, and is 30-40 mm. in length. 

The forms about to be described are common at Millport, and may often be found 
under stones between tide levels. When extended they are filiform, and from two to 
three inches (50-75 mm.) in length ; the proboscidean apparatus extends about one- 
third to two-fifths of the length of the animal; the distance between brain and 
mouth is about three times that between anterior end and brain. Since the union of 



8 DR J. STEPHENSON ON 

these forms into one, or their separation into two, species depends mainly on the facts 
of pigmentation, a full account of this character and its variations will be given. 

The general colour of most of my specimens was yellow ; in some cases it was a pale 
orange, deeper in tint in the anterior part of the body. In all cases the anterior tip 
of the head was brilliantly coloured, either a bright orange or a bright red ; this area of 
colour was not definitely limited, but was smaller in extent than that shown in 
Macintosh's pi. iv., fig. 5. The red pigment appeared to be partly diffuse and partly 
or mostly granular, and the individual grains had a bluish tinge ; two larger aggrega- 
tions, symmetrically placed one on each side, near the anterior end and near, but not 
at, the margin of the body, resembled two eyes ; a somewhat smaller and more 
anteriorly placed median spot was also sometimes distinguishable ; and, by focussing, 
it was found in one instance that the lateral spots were dorsal, the median spot, how- 
ever, ventral. 

In addition to the above general account, two observations may be given in detail ; 
both the following specimens were obtained from Fairlie, on the mainland. 

(a) In one of these specimens the pigment in the head consisted of (i.) an orange 
pigment, diffused (i.e. not aggregated into visible granules) over the anterior end of 
the head, continuous in area, not as two patches ; (ii.) a granular pinkish-red, with a 
slight blue tinge, at the tip of the head, apparently between the epithelium and the 
deeper layers, much more limited in extent than the orange ; (iii.) two large dark spots, 
of a dusky purple colour, one on each side near the anterior tip, but no median spot. 

(b) The second specimen contained ripe ova ; the whole anterior part of the body 
was more orange than usual, and this was especially marked over the anterior part of 
the head and over the oesophageal region. In this case the distinctive pigment at the 
anterior tip was mainly concentrated into two red patches, with an especially brilliant red 
spot in each ; there were no blue eye-like spots at all. There were also over the general 
surface of the body, but not of the head, a large number of minute white specks. 

It follows from the above descriptions taken together that forms are met with in 
colour from whitish through cream to yellow and pale orange ; that the anterior end of 
the body is in many specimens more deeply pigmented than the rest ; and that in a 
whole group of forms the extreme tip of the snout has a brilliant red or orange colour. 

Again, in the specimens with red tip, the red colour may be more or less extensive 
and more or less definitely localised. MacIntosh evidently considers it as an intensi- 
fication of the body-pigment, and shows it as of relatively considerable extent ; Joubin 
also indicates that it is not delimited from the general red colour of the head ; my own 
specimens show a more limited extent of brilliant red than Macintosh's, but equally 
with his and Joubin's it is not sharply defined ; while Burger, on the other hand, finds 
two very small bright red spots. 

Further, in these specimens with red tip, the pigment may be in part granular, and 
the granules may have a bluish tinge. Aggregations of these granules may be dis- 
tinguishable to the number of two or sometimes three ; such aggregations look like 



THE NEMEETTNES OF MILLPORT AND ITS VICINITY. 9 

eyes, and may appear bluish-red, dusky purple, or even black. Macintosh apparently 
did not meet with these forms. 

MacIntosh gives the length of his specimens as three to four inches (approximately 
75-100 mm.) ; Joubin gives 600 mm. as the maximum for C. linearis, and says that 
C. bioculata is shorter than C. linearis; Burger makes C. linearis 100-150 mm., and 
C. rufifrons ( = bioculata) only 30-40 mm. ; my specimens were 50-75 mm., on the whole 
somewhat shorter than Macintosh's and intermediate between Burger's two forms. 

If now it is decided to separate these forms into two species, this may conceivably 
be done either on the basis of the presence or absence of a red tip to the snout, or on the 
aggregation or otherwise of the bluish granules into eyelike masses. The presence of the 
first of these characters is the origin of the appellation rufifrons, the second of bioculata. 

The latter of the indicated alternatives is the one chosen by Burger in the 
Tierreich. The diagnosis of C. linearis runs : " Weiss offers mit gelblichem Anfluge. 
Ohne Pigmentflecke. L. 100-150 (nach Joubin bis 600); Br. 0*5-1 mm." That of 
C. rufifrons is : " Weisslich. Mit 2 kleinen rot-blauen Pigmentflecken an der Kopf- 
spitze. L. 30-40, Br. 0"5 mm." 

On this it may be remarked that the reddish-blue spots are variable features, 
consisting as they do apparently of closer or looser aggregations of scattered granules. 
Their size seems to vary, and also their colour, the latter feature probably according to 
the closeness of the aggregation ; their position, too, is not always the same, for while 
Joubin describes them as on the extreme margin, I have always found them a little 
distance within the margin ; their number also varies — sometimes three, sometimes only 
two. Again, if this feature be adopted as the criterion, the Millport specimen (b) above 
(which for the rest comes nearest of any of my specimeDS to Burger's Naples specimens 
of C. bioculata with two small bright red spots) would have to be excluded from C. 
rufifrons as showing no trace of a blue tinge ; while all my others, though with a less 
concentrated pigmentation, would nevertheless be C rufifrons. 

Burger himself gives us an example of the confusion w T hich results from taking 
this character as a criterion. As we have seen, he considers his Naples specimens (C. 
bioculata, = C. rufifrons of the Tierreich) to be the same form as Macintosh's specimens 
with red pigmentation. MacIntosh groups all his forms together as C. linearis, and 
accordingly we find in the Tierreich, as part of the synonymy of C. rufifrons, 
" C. linearis (part.) MacIntosh." It may, however, be inferred with certainty that 
Macintosh's specimens had no red-blue pigment spots, since they could not possibly 
have escaped his observation if present (he states, indeed, that there are no eyes) ; yet 
this is the character which to Burger is distinctive of C. rufifrons, under which he 
subsumes these specimens of Macintosh's. 

If the presence or absence of red-blue pigment spots fails as a criterion of distinction, 
much the same kind of objections seem to apply to the presence or absence of a red tip 
to the snout. The extent of pigmentation may be greater (MacIntosh) or less (Mill- 
port forms, Naples forms) ; it may be delimited in two patches (Burger), or continuous 

TRANS. ROY. SOC. EDIN., VOL. XL VIII. PART I. (NO. 1). 2 



10 DR J. STEPHENSON ON 

in area and not sharply defined (other authors). Macintosh's description seems to 
imply that he had had intermediate forms between the typical linearis and rufifrons 
forms before him, and he states that he does not consider the red pigmentation to be of 
specific import. Both he and Joubin evidently consider it to be an intensification of 
the general colour of the head. 

The impression given by a consideration of all the above descriptions is, I think, 
that of a continuous series, somewhat as follows. In the first place, we have forms of a 
pale cream colour without any special pigment in the head ; next, there may be a 
yellowish patch on the snout and a yellowish tinge in the oesophageal region ; then the 
tip of the snout may show special pigmentation — a deeper yellow, orange, or red — but 
the pigment is here not apparently granular in form. Again, while the general colour 
of the body is yellow or orange, the whole anterior part may be deeper in tint, the 
head being red, and the anterior tip most brilliantly coloured of all, but there is no 
sharp limitation of the red area ; the pigment is partly diffuse, but also partly granular. 
As the next stage in concentration, the red pigment may be mainly visible as two red 
patches, with, it may be, again a particularly brilliant red spot in each. 

This condition may be reached without the appearance of any blue tinge. As a rule, 
however, a blue tinge makes its appearance already in specimens with a pigmentation 
more diffused than that last described ; it may occur as a mere tinge in the red, or may 
give rise to the appearance of eye-spots by the aggregation of bluish-red pigment grains. 
Such spots may have a colour which, like that of the individual granules, may be 
described as bluish red, or a closer aggregation of the granules may result in a dusky 
purple or even a black. Of these ' eye-spots ' there may be either two or three ; if two, 
they are symmetrical, one on each side ; the third, when present, is median and anterior. 

The length of the Millport specimens, intermediate between the figures given by 
Burger for his two forms linearis and rujifrons, also tends to obliterate the distinction 
between them. 

For the above reasons, I believe that the names rujifrons and bioculata ought to 
disappear as specific appellations, and that the peculiarities to which these names refer 
constitute merely colour varieties of C. linearis. 

On the Physiology of the Circulation in Cephalothrix linearis. 

My observations on this subject are in the main confirmatory of those of MacIntosh 
on Cephalothrix (quoted below). In view of the little that is known concerning the 
physiology of the circulation in Nemertines, I may be permitted to call attention again 
to this form, in which the vascular phenomena appear to be particularly obvious. 

Under these circumstances it may perhaps be useful first to bring together, as far as has 
been possible to me, what has been written on the subject of the circulation of Nemertines. 
How contradictory, as well as scanty, are the statements concerning the physiology, as 
distinct from the anatomy, of this system, will appear from the following references. 

The most definite accounts are to be found in MacIntosh (10). With regard to the 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 11 

Enopla generally, he says : " The course of the circulation, so far as observed, is as 
follows. Posteriorly a gentle contraction from behind forward drives the contained 
fluid along the great central vessel to the front, where it is forced through the 
anastomotic into the lateral vessels and the cephalic arch. Each lateral trunk swells 
with the wave, and the fluid then proceeds to the posterior end to enter the median, as 
before mentioned. In addition to the stream poured into the lateral trunks, another 
passes into the cephalic arch by the vessel on each side, and the counter-currents must 
meet and commingle, returning again during the diastole of the central vessel." The 
contraction of the lateral trunks in Nemertes carcinopliila is stated to be very vigorous. 

With regard to the Anopla, he states that " the current is driven by the contraction 
of the vessels now backward, now forward, so that it is rather a kind of oscillation. . . . 
The dorsal generally contracts from behind forward, and drives the corpuscular fluid, 
not only to the front, but also through the transverse branches into the lateral trunks. 
The latter propel their contents in both directions." And with regard to Cephalothrix, 
" in the living animal each lateral vessel contracts regularly and swiftly from before 
backwards, sending a wave of fluid towards its posterior end, at which the contraction 
ceases. A reversed movement by-and-by takes place, the contents being propelled 
towards the snout. . . . There appears to be little regularity or rhythm in the 
movement of the fluid in these vessels, both occasionally contracting from before 
backwards at the same time. Generally, however, the contractions are alternate." 

Oudemans (11), who made a careful study of the circulatory system in a number of 
forms, seems to have worked entirely with preserved material, by the method of serial 
sections. 

Joubin (8), says : " Les vaisseaux .... sont animes de contractions et de batte- 
ments, mais il n'y a point de coeur distinct. Le sang progresse vers la tete dans les troncs 
lateraux, vers la queue dans le tronc median." Speaking of Tetrastemma Jiavidum, he 
says : " Les vaisseaux sont remplis de sang rouge et font des ondulations colorees, saillantes 
sous la peau." The same author, in a subsequent work (9), says : " Le sang circule 
grace a la contractilite des vaisseaux, car il n'y a aucun organe central de propulsion ; il 
suit une direction determines ; vers la tete dans le vaisseau median, vers la region caudale 
dans les vaisseaux lateraux " (thus agreeing with MacIntosh, and reversing his previous 
statement). " Le sens contraire est indique par Vogt et Yung pour le Tetrastemma 
Jiavidum. Quand on observe a letat vivant certaines especes transparentes, on voit 
les ondes contractiles progresser lentement sur les vaisseaux a intervalles reguliers." 

Burger (2) says, with regard to the Anopla : " Das Blutgefasssystem wird von mit 
Muskeln ausgestalteten Stammen gebildet, welche eine Fliissigkeit, die freie Zellkorper 
enthalt, durch den Korper pulsiren lassen ; " but I cannot find any reference to the 
Enopla. I find nothing concerning the physiology of the circulation in the same author's 
great monograph (3), nor in his later contribution to Bronn's series (4) ; though Benham 
(1) says that, according to Burger, "the blood, in Metanemertines, flows out of the 
dorsal, through the circular vessels, into the lateral ones, returning to the dorsal vessel 



12 DK J. STEPHENSON ON 

at each end of the worm," which is practically Macintosh's statement ; but he adds 
that this is very unsatisfactory and uncertain. 

Punnett, in his Addenda and Corrigenda to this volume of Lankester's Treatise (13), 
says that " in spite of what has often been written to the contrary, it is exceedingly prob- 
able that in most cases, if not all, the blood vessels are destitute of muscle fibrils; and that 
the blood is kept in circulation by the waves of contraction passing over the body-wall." 

Finally Coe (6), in describing Carinomella, states that " shortly behind the nephridial 
region the lateral vessels acquire muscular walls and are strongly contracted at intervals." 
He does not describe vascular contractions in any other forms ; and in the general section 
of his work, under " Histological Structure," after describing the muscular coat of the 
chief vessels, says: "The transverse vessels and many of the lacunge are without 
muscular walls, and even where muscles occur, the circulation of the blood is dependent 
mainly on the movements of the body as a whole, and as a rule passes backward and for- 
ward irregularly in any of the vessels." 

The circulation in the lateral vessels of C. linearis can be seen in specimens 
compressed under a coverslip. The phenomenon appears as a series of waves, passing 
along in the situation of the vessels, laterally to the alimentary canal. The walls of the 
vessels are not themselves visible, and the waves manifest themselves as the rapid 
passage of a lighter streak or patch along the body of the animal. 

The rhythm is irregular. The waves pass in all cases through the greater part of the 
length of the animal, it may be from the region of the nerve ganglia anteriorly to the 
extreme posterior end. The passage of the waves was always rapid. 

In direction the waves may be either postero-anterior or antero-posterior ; and 
these usually alternate, with, it may be, a second's interval between the arrival of the 
wave at the head, and the starting thence of a wave in a posterior direction. The waves 
on the two sides may be almost synchronous ; but I have observed a wave on one side 
only, quite unaccompanied by any indication of a wave on the opposite side. 

The specimens were, as a rule, motionless during the observation of the phenomenon. 

With regard to Cephalothrix, then, it seems allowable to conclude : — 

(i. ) That there is a definite circulation in the lateral vessels, occasioned by a series of 
contractile waves alternately postero-anterior and antero-posterior. 

(ii.) That this is not due to contractions of the body-wall, nor to movements of the 
body as a whole. 

(iii.) And that it would therefore seem necessary to assume the presence of muscular 
tissue in the walls of the vessels. 

Emplectonema gracile (Johnst.) ( = Nemertes gyxicilis). 

Near low-water mark, under stones, Balloch. Not uncommon. 

Length up to seven inches ; proportionally very thin ; not markedly pointed at the 
posterior end. 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 13 

The general colour is brown, with sometimes a greenish tinge, especially towards and 
on the head ; the green colour can sometimes, especially from the ventral surface, be 
seen to be due to the diverticula of the alimentary canal. In one specimen, the green 
became a distinct blue on the anterior part of the head. Sections show a blue granular 
deposit in the walls of the alimentary canal, especially in the anterior part of the body. 
The ventral surface is lighter in colour ; the margins of the body are clearer ; a thin 
stripe of a lighter tint in the middle line of the anterior part of the body is due to the 
proboscis cavity. The genital products may show as yellowish masses within the lighter 
margins of the body. 

The head is somewhat circular in shape, with median anterior notch ; it is broader 
than the succeeding part of the body, and is fairly well marked off. The eyes are 
numerous, and not usually distinctly arranged in two groups on each side. No grooves 
were to be seen in the living animal, though they are apparent in transverse sections of 
the head. 

The stylet is very large, with a sabre curve, and is situated very near the anterior 
end of the body ; the basis (v. fig. 6) is much elongated, at least twice as long as the 
stylet, and swollen at its proximal end. There are two reserve sacs ; seven stylets were 
counted in each. 

The condition as to head-glands and other gland-cells in the head, the cerebral 
organs, and position of the mouth are as given by Burger (3, 5). The statement in the 
Tierreich, that " der Blinddarm reicht bis in die Nahe des Gehirns," is a little mis- 
leading ; it is, as in E. neesii, two long anteriorly directed diverticula of the cgecum, one 
on each side, which nearly reach the brain, and the condition might have been described 
in terms identical with those used in the definition of the latter form. 

The differences between the above description and the accounts of previous observers 
are, except perhaps with regard to the eyes, not considerable. Green or brownish 
(greyish, olive-) green appears to be the commonest colour ; Macintosh and Joubin 
have noted a bluish tinge. Macintosh's plate does not show the head as being marked 
off from the body. 

The eyes are stated (Macintosh, Burger) to be arranged in two or three groups 
on each side. Macintosh adds that the middle group on each side is nearer the middle 
line of the head ; his plate, on the other hand, does not show any marked division into 
groups, nor is the middle group, such as there is, any nearer the middle line than the 
anterior group — rather the reverse, in fact. Burger (3) adds that the eyes are in 
rows in the anterior, massed together in the posterior group. 

Emplectonema (Nemertes) neesii (Orst.). 

Common, especially in the byssus of mussels, on the Keppel pier ; also between 
tide levels. 

Length two to four feet (considerably greater than the lengths found by previous 



14 DR J. STEPHENSON ON 

observers) ; breadth 4 mm. Slow-moving ; placed in a vessel of water, slowly uncoils 
and extends itself as a long strip round the sides of the vessel at the level of the surface 
of the water ; the tail tends to curl in a corkscrew fashion. 

Colour of the dorsal surface brown, with often a tinge of purple ; on examination 
with a lens, the colour appears as a dark brown mottling ou a lighter ground. Head 
slightly lighter in tint ; under surface light greyish, with darker bands one on each 
side of the middle line ; the intestine is seen on the ventral surface as a pale yellow 
line giving off branching diverticula. 

Head not expanded, not marked off from the body ; tail slightly tapering. 

Cephalic grooves small, oblique, on the under surface of the head near the tip, 
approaching one another at their anterior ends. 

Eyes very numerous, small, in two groups on each side, a more numerous anterior 
and less numerous posterior ; on focussing, they appear some distance beneath the 
surface, and the examination of sections shows them to be beneath the epithelial 
layer. 

The head glands are well developed, and there are also, as noted in the specific 
diagnosis in the Tierreich, a large number of subepithelial gland-cells in the head. It 
may be added that gland-cells extend continuously for some distance along the body, 
aggregated in two rows, one along each side ; the masses are very conspicuous to the 
naked eye in a series of sections, since they stain deeply (with hsematoxylin), and are of 
considerable size ; they displace the longitudinal muscle layer, and impinge on or even 
surround the nerve cord. The gland-cells of the head, as they are traced backwards in 
serial sections, leave first the mid-dorsal area, then the mid-ventral, and so come to be 
aggregated in the two lateral rows described above. 

The cerebral organs and the anterior diverticula of the alimentary canal agree with 
previous descriptions. I have not, however, seen any statement as to the number of 
proboscis nerves ; there are definitely twelve in one of my specimens ; but the number 
of nervous tracts differentiated in the nervous sheath perhaps varies, since in another 
specimen the nerves themselves appear to be less definite, and their number seems to be 
greater (fourteen or sixteen). 

With regard to the coloration of this form, there is a general agreement in the 
descriptions to which I have been able to refer; and to these my specimens also 
conform. The descriptions of the cephalic grooves by MacIntosh and Joubin differ 
slightly, and the account given above accords rather with that of MacIntosh. The 
same may be said with regard to the eyes ; Joubin figures them as aggregated on each 
side into a large group of curious shape, extending farther back than in the Scotch 
examples, and not separated into two groups on each side. The length of my specimens 
(2 to 4 feet) is to be compared with the figures given by MacIntosh (4 to 18 inches) 
and by Burger in the Tierreich (up to 460 mm. = 18 inches), as well as by Joubin 
(8) (50-60 cm. = 20-25 inches). 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 15 

Amphiporus lactifioreus (Johnst.). 

Fairly common at and near Millport. 

Length, f-5 inches ; breadth, average, 1-lJ mm. ; body dorso-ventrally flattened ; 
head slightly expanded, frequently somewhat diamond-shaped with a blunt point 
anteriorly ; the hinder end does not taper to a point. 

The animals contract very markedly on interference, and assume a slug-like shape ; 
a specimen an inch and a half long will shorten to a third of an inch. The surface 
secretes a very sticky mucus. Progression may take place by an equable gliding, or by 
the passage over the body of a series of contractile waves ; or the head may crawl 
evenly while the tail executes swimming movements. 

Two varieties of colour were met with ; the first includes specimens from a light 
purple to a delicate French grey ; the second includes forms which may be described as 
cream, yellow, pale orange, or pale flesh colour. The posterior part of the body is often 
darker than the anterior ; there is a white line, due to the proboscis, down the middle 
of the dorsum, and the margins of the animal are paler or more translucent. 

The cephalic grooves are in two pairs, an anterior and a posterior. The anterior 
notch the margin of the head at its widest part, and are continued thence somewhat 
backwards on the dorsal, obliquely forwards on the ventral surface. The posterior pass 
obliquely backwards on the dorsal, transversely on the ventral surface, and nearly meet 
those of the opposite side in the middle line both dorsally and ventrally. 

The above description of the grooves agrees in the main with that of Macintosh, 
who, however, states that the anterior grooves run forwards on the dorsal surface, not 
backwards. Joubin (8), on the contrary, shows quite a different form for the posterior 
grooves, which, in his specimens, had on both dorsal and ventral surfaces something 
the shape of the letter M. 

The eyes are usually, not always, in two groups on each side, which are divided 
by the anterior furrows ; I have never seen three groups. The number of the eyes 
varies very much ; the smallest numbers I have noted are three in each anterior, 
two in each posterior group, or ten in all ; but twenty -four, twenty-seven, thirty-four, 
and thirty-six are met with. The largest eyes are in the posterior group. It may be 
mentioned that Burger's Naples specimens possessed a very large number of eyes, 
twenty in each of four groups. 

Another, and a more important, variation between the Naples and the Millport 
specimens has to do with the shape of the basis of the stylet. This is shown by 
Burger as being markedly constricted a little posterior to the middle of its length ; 
its thickness is about equal in front of and behind the constriction ; since, however, 
according to the figure, the constriction is nearer the posterior end, the anterior portion 
of the basis is the bulkier. A similar description (" plumpe Sockel, in der Mitte 
ringsum eingeschnurt, vorn und hinten fast gleich dick") is repeated in the specific 
diagnosis given in the Tierreich. 



16 DR J. STEPHENSON ON 

The shape of the basis in my specimens is seen in fig. 7. It will be seen that 
the constriction is very slight, if indeed it exists — if, that is, the appearance of a 
slight constriction be not merely due to the rather greater width of the posterior 
rounded end of the basis ; the posterior part of the basis is, owing to this greater 
width, bulkier than the anterior. Fig. 8 represents a different shape, met with on 
one occasion only ; the basis is here much shorter and broader than usual, and resembles 
a truncated cone. In the two reserve sacs I have found either two or three stylets. 

The number of proboscis nerves is not given in any account of the species that 
I have seen. Since it varies in the different species of this genus, and since it is a 
character that is sometimes used as a specific distinction, it may be worth while stating 
that the number is fourteen. 

The csBcuni sends off a pair of long and slender prolongations, which reach as far 
as the brain. These, which are usually described as hollow pouches, are in my prepara- 
tions solid throughout. As stated by other authors, the cerebral organs are large, and 
situated in front of the brain. 

This form appears to be a good example of the variability of the class. Joubin's 
account of the cephalic grooves differs considerably from those of MacIntosh and 
myself; Burger's specimens differed from the British forms in being found at some 
depth, in possessing a much larger number of eyes, in the shape of the basis of the 
stylet, and in coiling themselves up, not contracting themselves slug-like ; the colour 
also, as in most members of the class, is variable; and even in the Millport forms I 
found in one instance, as has been noted, a considerable divergence from the rest 
in the shape of the basis of the stylet. 

Amphiporvs pulcher (Johnst.). 

This species has been the subject of a number of descriptions which vary from each 
other considerably in certain points, e.g. the eyes, the cephalic grooves, and the stylet. 
A comparison of the forms met with at Millport with those described from other 
localities may therefore be of interest. 

Specimens are rare at Millport ; three were dredged in fifteen fathoms off Ascog 
Bank. This supports Joubin's statement (8) that "on ne trouve pas non plus les deux 
especes {i.e. A. pulcher and A. lactijloreus) dans les memes localites." 

The length was about 1|- inch, the body relatively broad and flat ; the tail much 
flattened ; the head not wider than the rest of the body, somewhat rhomboidal in shape, 
ending anteriorly in a blunt point. The animals are sluggish in habit, contracting 
readily to a slug-like mass ^ of an inch long ; they can swim lazily, ventral side upper- 
most, at the surface of the water. I have not seen the peculiarity noted by MacIntosh, 
that, when irritated, they turn on edge and swim rapidly through the water by swift 
lateral strokes of the oar-like extremity. 

The colour is a light orange-pink, the margins and ventral surface being lighter, of a 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 17 

very light flesh colour. The proboscis, readily extruded, is also of a delicate pink 
colour. The animals very readily break up, more readily than any other Nemertines 
met with at Millport. 

The cephalic grooves are two on each side, an anterior and a posterior ; and each 
groove extends on both dorsal and ventral surfaces (v. figs. 9,10). Dorsally the anterior 
grooves run transversely, nearly meeting in the middle line, and giving off a number of 
small secondary grooves which run in an anterior direction at right angles to the main 
groove ; the posterior grooves run very obliquely backwards and meet in the middle 
line. Ventrally, the anterior grooves are continued transversely, meeting or not in the 
middle line, with secondary grooves of the same description as those on the dorsal 
surface, and a dimple nearly half-way between the lateral margin and the median line ; 
the posterior grooves run transversely, with a slight inclination forwards, but do not 
meet. 

The above description corresponds in most points with that of Macintosh, except 
that this author neither figures nor describes any extension of the posterior grooves on 
the ventral surface. Joubin (7) criticises Macintosh (" la figure donne par Macintosh 
est assez defectueuse pour ce qui est des sillons "), and states (8) " je n'ai point vu les 
sillons secondaires que figure Macintosh." He describes (8) the anterior dorsal grooves 
as being together V-shaped, like the posterior, but his figure shows that they run for the 
most part almost transversely, each, however, bending in a posterior direction near its 
inner end ; the ventral grooves of his specimens seem to have been entirely different 
from the Scotch forms, and, according to his figure, produce by their bifurcations and 
reunions the appearance of a number of somewhat rhomboidal or irregular areas on this 
surface. Burger (3, 5), as for A. lactifloreus, does not mention the grooves. It would 
seem therefore, on the whole, that both the course of the main grooves and the presence 
of secondary grooves are liable to very considerable variations. 

The eyes in my specimens were numerous, and distributed somewhat irregularly 
near the lateral margins of the head, mostly in front of the anterior groove. In one 
specimen there were thirty -four on the left side and about as many on the right ; in 
another, there were fifteen on each side. The larger eyes are irregular aggregations of 
pigment, apparently formed- by the fusion of smaller eye-spots. 

Other authors give diverse accounts of the eyes. MacIntosh states that they are 
about twenty-three in all ; Joubin (7) that they are thirty-five to forty-five on each side, 
or, again (8), eighteen to twenty-five on each side. Burger (3) states that they are 
" ziemlich regelmassig zweireihig angeordnet," and, in the Tierreich, includes this feature 
in the diagnosis of the species (" mit vielen Augen, die in 2 Reihen angeordnet sind"). 
From what has been said, however, it would appear that neither the number nor the 
arrangement of the eyes present sufficient constancy to permit of their being used as a 
specific distinction. 

The basis of the stylet is of a regular oval shape, and the stylet itself may project 

behind it (fig. 11). The distal free portion of the stylet is equal in length to the basis. 
TRANS. ROY. SOC. EDIN, VOL. XLVIII. PART I. (NO. 1). 3 



18 DR J. STEPHENSON ON 

There are two reserve sacs, the numbers of reserve stylets found in these were six, 
seven, eight, or nine. 

Here, again, Macintosh's figure (his pi. xii., fig. 6) shows a close agreement with 
the above description. Burger, however, appears to have been dealing with a very 
different shape of pedestal ; compare the latter author's plate (3), and also his statement 
that the basis " ist kegelformig und hinten etwas kuglig angeschwollen." A similar 
statement is made part of the specific description in the Tierreich. Students of the 
Nemertini will for ever be indebted to Burger's comprehensive and masterly work ; 
but if a criticism may with all respect be ventured, I would suggest that he has relied 
too much on his own specimens for the construction of specific diagnoses, and has not 
allowed enough weight to the descriptions of previous authors, nor considered sufficiently 
the great variability of many species of the class. 

With regard to certain other specific characters given in the Tierreich, the ccecum 
in my specimens corresponds to what is there stated ; the cerebral organs, stated to lie 
behind the brain, I find to be for the most part alongside and almost coextensive antero- 
posteriorly with the brain, though they extend backwards for some little distance be- 
hind it ; the number of proboscis nerves, however, is different — Burger giving ten, 
while I find twelve. 

Amphiporus elongatus, n. sp. (fig. 12). 

A single specimen ; found on Fairlie sands. 

Length, 3 inches ; filiform, breadth when extended being less than 1 mm. ; tapering 
markedly towards the head, which is flattened ; the tail blunter than the head. 

Colour bright yellow, becoming an orange yellow when contracted ; the margins 
lighter, the ventral surface of the same colour as the dorsal. 

The head was flattened, tapering, and not marked off from the body ; the ganglia 
were visible as reddish masses. The eyes were five in number (v. fig. 1 3) — two smaller 
ones, one on each side near the tip of the snout, and three larger ones, two on one side, 
and one on the other, over the cerebral organs ; there was a considerable interval 
between the two sets. 

The cephalic grooves were two on each side (v. fig. 13), the anterior very oblique, 
leading backwards on the dorsal surface from the lateral margin to the cerebral organ ; 
the posterior also leading obliquely backwards almost parallel to the anterior pair, and 
nearly meeting in the middle line at the level of the anterior end of the brain. Both 
sets of grooves were continued round the margins on to the ventral surface in the same 
direction, i.e. running from the lateral margin forwards and inwards towards the middle 
line. 

The cerebral organs were large, and situated quite in front of the brain (v. fig. 13). 

The 'proboscis sheath was continued to within a very short distance of the hinder 
end. The stylet was rather thick as compared with its length, and shorter than its 
pedestal ; the shape of the latter was roughly cylindrical, with rounded ends, very 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 19 

slightly constricted nearer the proximal (base) than the distal end (v. fig. 14). There 
were two reserve sacs, containing respectively one and two stylets. The rhynchocoelom 
had no diverticula. 

The alimentary canal possessed a large number of lateral diverticula, attached by 
strands to the body-wall ; a diverticulum was sometimes attached by two strands, a 
slight bifurcation of its extremity being thus produced. When the whole animal was 
examined under the microscope, the diverticula appeared as a series of leaf-like appen- 
dages of the alimentary tract, i.e. were flattened from before back. 

Examined in spirit several months afterwards, the specimen was about an inch and 
a half long, a millimetre broad, and of a light brown colour ; tapering gradually at one 
end, more suddenly at the other. 

Examination of Sections. — Portions of the animal were sectioned for further 
investigation of certain characters used in discriminating species of this genus. The 
following additional features may be noted : — 

The ventral ganglia of the brain are much thicker than the dorsal. There are no 
neurochord cells or neurochord. The sections confirm the above statements as to size 
and position of the cerebral organs. The caeca of the alimentary canal can be seen, 
in horizontal sections, sometimes to be bifurcated ; the anterior ventral caecum is 
of large size, and sends forwards a pair of diverticula, which do not nearly reach the 
brain. Testes are scattered between the alimentary diverticula but not as a regular 
series alternating with these latter. The (presumably mucous) contents of many of the 
epidermal cells are intensely stained (by Delafield's haematoxylin). The head-glands 
are not strongly developed. The proboscis, as seen in transverse sections, lies in a 
spacious rhynchocoelom, but is itself of very small diameter, and the proboscis nerves 
are not to be made out. 

Since this form differs from most species of the genus Amphiporus— which are, as a 
rule, somewhat compact and thickset forms— in being thin, indeed filiform when extended, 
I have chosen for it the specific name elongatus. 

I propose the following diagnosis for this form : — Length 75 mm., breadth ivhen 
extended less than 1 mm., filiform. Head tapering, not marked off from body; 
tail blunter. Colour bright yellow, the same on both dorsal and ventral surfaces; 
margins whiter. Eyes few in number, in two groups on each side. Stylet shorter 
than its pedestal ; the latter roughly cylindrical, slightly constricted posterior to its 
middle ; two reserve sacs. Head grooves tivo on each side, directed obliquely bach- 
wards and inivards on dorsal surface, continued over margin of head on to ventral 
surface. Cerebral organ in front of brain. 

Prostoma (Tetrastemma) candidum (Mull.), 
This worm is commonly found at and near Millport, under stones between tide-marks. 
Previous authorities have differed considerably in the characters which they attribute 
to the species ; the following account may therefore be of interest. 



20 DR J. STEPHENSON ON 

In length most specimens were from one-third to two-thirds of an inch, some few 
longer, some even shorter ; they agreed therefore rather with Burger's Naples 
specimens (scarcely more than 1 cm.) than with Macintosh's (1-1^ inches). In 
breadth they were about 1 mm. They were thus relatively short, and sausage-shaped ; 
the posterior end was narrower and somewhat pointed. The animals are active, can 
swim freely, and are very hardy and resistent under examination. 

Green and yellow are noted by previous authors as being the usual colours ; Joubin 
(8) adds white and red. In my specimens the range of colour was extensive— yellow, 
light orange, light greyish yellow, light brownish grey, light brown ; one specimen was 
noted as being between pink and brick-red, the colour being due to the intestine. A 
green specimen, almost certainly of this species, was given me one evening by Dr A. L. 
King. I was just leaving the laboratory, and deferred a detailed examination ; but the 
animal, which was damaged, had disintegrated before the next day. The head, or the 
tail, or both, are lighter in tint, and the margins of both head and body are more trans- 
lucent. The ova may appear as very marked yellow masses in a row on each side of the 
middle line. 

Certain markings on the head deserve notice. MacIntosh has described a pale 
streak in the median line anteriorly ; and has noted that in some individuals a few 
white grains are to be seen between the anterior pair of eyes. Burger, however (3), 
found no pigment spots or streaks on the head. In the Millport specimens there may 
be a pale streak in the middle line in front of the eyes ; certain specimens, again, show a 
pair of white splashes, each a collection of minute white specks, between and in front 
of the anterior pair of eyes ; while others show a pair of dark pigmented patches, 
elongated in shape, broader anteriorly, one on each side of the dorsal surface of the 
head, stretching from the anterior to the posterior eye ; but none of these markings on 
the head were constant. 

The shape of the head is much used in this genus as a means of discriminating the 
various species. MacIntosh describes it as being wider than the succeeding part of the 
body, Burger as being broadened and of an elongated egg shape. I have found it as 
a rule a somewhat short oval, sometimes rather flattened in front, occasionally shorter 
and almost circular in shape ; it is slightly — sometimes very slightly — broader than 
the succeeding part of the body, from which it is marked off by a constriction ; there is 
a slight median notch with well-marked cilia in the middle line anteriorly. 

The eyes, in this genus four in number, are, according to MacIntosh, set in a square ; 
according to Burger, they are very small, and are set in a rectangle. In my speci- 
mens they were arranged in a rectangle, the side lines of which were as a rule distinctly 
longer than the front and back ; occasionally, in an animal with shorter and more 
circular head, they formed a square. They were, for the size of the animal, large and 
conspicuous ; and in one specimen the anterior pair were rather larger than the 
posterior. 

The cephalic grooves are one pair, which notch the margins of the head in front of 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 21 

the level of the posterior eyes ; thence they incline obliquely backwards on the dorsal 
surface, passing just in front of the eyes and nearly meeting in the middle line ; on the 
ventral surface they are continued obliquely forwards for some distance from the 
margin. 

The basis of the stylet is, according to Burger, constricted, the constriction being 
situated, according to the figure (3), nearer the posterior end ; the anterior swelling is 
as thick as the posterior ; the basis is longer than the stylet. I find the basis elongated 
in shape, rounded and expanded posteriorly (v. fig. 15), and slightly longer than the 
projecting part of the stylet. There may be an appearance of a slight constriction 
around the middle of the basis, but this is due to the bulging of the posterior end ; and 
an actual constriction does not exist, or is of the very slightest, the sides of the 
anterior part of the basis being parallel to each other. 

There are two reserve sacs, with two, three, or four reserve stylets in each ; in one 
case there was only one reserve stylet on one side, two on the other. Burger, at 
Naples, found two ; in the Tierreich he gives two or three as the number. 

That this is the form commonly known as Tetrastemma candidum there can, I think, 
be no doubt. The only form with which it could be confused would be the nearly 
related Prostoma {Tetrastemma) jiavidum, but the colours are distinctive — yellow and 
green for candidum, pink for Jlavidum. Again, the head is wider than the body in 
candidum, as generally in my specimens ; and the white grains between the anterior 
eyes, described by MacIntosh, were also seen in some of the Millport forms, 

It may be noticed here, again, how considerably Burger's Naples specimens diverge 
from the description here given. In the length, in the broadening of the head, and in 
the position of the cephalic groove on the dorsum, the two correspond ; and in colour 
Burger's specimens — of a light or dark green, with yellow margins and yellow head 
— agree in a general way with Macintosh's description, if not with mine. But, on the 
other hand, Burger found no pigment spots or streaks on the head, the eyes were " very 
small," and his description of the basis of the pedestal is quite different from that 
given above. In the Tierreich we find certain of these variations given as parts of the 
specific diagnosis; thus, the length is limited to 10-12 mm. (though MacIntosh had 
previously given 1-lJ inches), the eyes are stated to be very small, the pedestal to be 
moderately constricted, and the number of reserve stylets is limited to two or three. 
A revision of the definition would therefore seem to be advisable. 

On the Distinction of P. candidum from P. jlavidum. 

If the descriptions of the above two forms, as given by MacIntosh, Joubin, and 
Burger in his Naples monograph, as well as the diagnoses of the latter author in the 
Tierreich, be compared, it will be found that a number of characters are given by each 
author which might serve to distinguish the two forms from each other. But, on a closer 
examination, it will appear that the points of difference are differently stated by the 



22 OR J. STEPHENSON ON 

various authors — that is, that there is no general agreement as to what the points of 
difference are. It will further appear from the above descriptions, that of the Millport 
specimens included, that these forms, or at any rate P. candidum, are subject to a great 
amount of variation. And it will be seen also that specimens have come under examina- 
tion which unite in themselves characters ascribed to the two forms separately. 

With regard to the first point, it appears from a tabulation of the characters of the 
two species, as given by the several authorities, that the only distinguishing character 
in which all agree is that in P. candidum the head is broader than, in P. Jlavidum of 
the same width as, the succeeding part of the body ; # even this is not stated directly 
by Burger with regard to Jlavidum, but it may perhaps be inferred from the fact that 
the head of Jlavidum is said to be not marked off from the body. 

Thus, with regard to colour, while it is generally agreed that green or yellow, or 
both, are the characteristic colours of candidum, and a pinkish or a rose colour of Jlavi- 
dum, Joubin finds red specimens of candidum also. Macintosh alone mentions various 
markings on the head of both species. Joubin states that the head is long in Jlavidum, 
and the distance between anterior and posterior eyes greater (" leurs ganglions sont tres 
allonges comme la tete en general, dont les yeux sont plus distants que dans aucun 
autre Tetrastemma ") ; Macintosh finds the eyes arranged in a square in candidum, 
while the interval between anterior and posterior eyes is greater in Jlavidum ; but 
Burger makes no distinction between the two forms as regards this feature. MacIntosh 
alone notes that candidum is of active, Jlavidum of a more sluggish habit. Differences 
in the relative length of stylet and basis, and in the shape of the basis, have been noted 
only by Burger. 

To the fact that there is no general agreement as to the distinguishing characters 
of the two forms is to be added the fact that the forms themselves are very variable. 
This may be illustrated by the different colours and combinations of colours assumed, 
especially by P. candidum ; Joubin specially notes the variability of this form, and 
gives for it a greater range of colour than other authors ; he mentions white, yellow, 
green, and red. The length, again, as given by Burger, is in the case of P. candidum 
only one-third of that given by MacIntosh, and little more in the case of P. Jlavidum,. 
The size of the eyes, according to Burger, is "very small" in both; this is not noted 
by other authors, and was very distinctly not the case in the Millport specimens. With 
regard to the shape of the basis of the stylet in P. candidum, Burger, as has been 
seen, gives a description which is quite inapplicable to my specimens ; and with regard 
to P. Jlavidum, while the figure in his Naples monograph shows the basis to be some- 
what dumbbell-shaped, with a constriction in the middle and rounded ends of equal 
size, his description of this structure in the Tierreich runs " der kegelformige, kaum in 
der Mitte eingeschniirte Sockel." And it lias been seen that in my own specimens 
of P. candidum, in addition to well-marked colour differences, the length, the pig- 

* Jouuin has not given full descriptions of these forms ; in points not specified by him lie associates himself with 
.MacIntosh, and accordingly is counted as agreeing with this author in such cases. 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 23 

mentation of the head, the occurrence of white splashes in front of the eyes, the number 
of stylets in the reserve sacs, the shape of the head, and the disposition of the eyes, 
were all found to be variable. 

Lastly, the Millport specimens in several cases showed a combination of the characters 
of P. candidum and P. Jiavidum, and seem thus to have been to some extent inter- 
mediate between the two. While agreeing on the whole more closely with P. candidum,, 
in one case a pinkish colour was found, pink or rose being the characteristic colour of 
jiavidum. The arrangement of the eyes in a square (characteristic, according to 
MacIntosh, of candidum) was as a rule replaced by the arrangement in a rectangle 
(characteristic of jiavidum). The head of my specimens was as a rule only slightly, 
sometimes very slightly, broader than the next succeeding part of the body — the 
condition, that is, was intermediate between those characteristic of candidum and 
jiavidum respectively. And while the anterior and posterior eyes were as a rule of the 
same size, in one instance the condition of P. jiavidum was reproduced, in which form, 
according to MacIntosh, the anterior pair of eyes are larger than the posterior. 

I do not consider that any of my specimens represented the typical form of 
P. jiavidum ; and it would therefore be unwise in me to pronounce a definite opinion 
on the following point. But having regard to the want of any general agreement as to 
the characters by which the above two forms are to be distinguished, and, in addition, 
to the great variability of the forms themselves, as well as to the fact that forms exist 
which may be considered as intermediate in character, it seems not improbable that it 
will be found advisable to unite the two under a common designation. 

Oerstedia dorsalis (Abildg.) ( = Tetrastemma dorsale). 

A single specimen was dredged in the channel near the Biological Station. 

Length one-third of an inch ; comparatively stout, cylindrical or sausage-shaped, 
of generally stiff appearance ; the blunt posterior end tapers very little. The specimen 
frequently doubled itself up so that its two halves lay apposed side by side. It pro- 
truded its proboscis on apparently no provocation ; it ultimately broke into two under 
examination, and the anterior part then vomited and shook off the proboscis. 

The general colour was yellow to orange, very slightly lighter on the ventral surface ; 
a brownish pigment was also present, with a double arrangement — (i.) in three bands, 
two lateral and one mid-dorsal, not very definite; (ii.) in transverse bands over the 
dorsum, which became fairly conspicuous when the animal contracted itself, and which 
then gave the animal an annulated appearance. Scattered over the body were a number 
of minute white dots, easily seen with a low power of the binocular microscope ; these 
were specially aggregated to form a mid-dorsal line, superposed on the median dorsal 
brown-pigmented streak. The animal frequently twisted itself so that the mid-dorsal 
white line appeared as a spiral round the body. 

The head was flattened, not broader than the body, not marked off, tapering some- 



24 DR J. STEPHENSON ON 

what towards the front. The proboscis, when extruded, was bulky and long. A pair of 
transverse grooves, which indented the lateral margins, were visible one on each side 
dorsally behind the posterior eyes. 

The eyes were four in number, large, reddish brown, apparently situated some distance 
beneath the surface. The distance between anterior and posterior eyes of the same side 
was considerably greater than that between the two anterior or the two posterior eyes. 

The stylet was of the same length as its basis ; the latter was somewhat conical in 
form, with rounded end (v. fig. 16). Here, again, I find a considerable difference from 
Burger's figure (3), which shows a somewhat dumbbell-shaped basis, markedly con- 
stricted, with rounded ends of about the same size. 

There were two reserve sacs, each with four stylets. 

Linens longissimus (Gunn.). 

Not uncommon ; under stones near low- water mark, Balloch. 

No great lengths were met with ; 10 feet is not uncommon ; breadth | to ^ inch, large 
specimens \ inch. 

The animals are sluggish in habit. The general colour is from a dark brown to a 
dark velvety black, with a slight purple iridescence ; the ventral surface is a little paler 
than the dorsal. This general ground colour is varied by a longitudinal striping of a 
lighter tint ; distinct lateral and mid- ventral lighter stripes are always present, but the 
dorsal stripes are variable — of these latter there may be several, running parallel. All 
the stripes are most marked in the anterior part of the body ; some of the larger speci- 
mens had none on the dorsal surface except a mid-dorsal stripe, and that only on the 
head. 

The shape of the head would seem to merit a short consideration. Macintosh 
describes the head simply as being wider than the succeeding portion of the body. 
Burger, however, lays more stress on this character, and writes, in the diagnosis of this 
form in the Tierreich : " Kopf verbreitert, spatelformig, nicht vom Rumpf abgesetzt ; " 
while in his key to the various species of the genus the distinguishing character is " Kopf 
auffallend stark verbreitert." Joubin (8) does not mention a broadening of the head, 
and his figures show it as slisfht or absent. 

In the Millport specimens the head has whitish margins and is somewhat flattened. 
When the animals are contracted, it is narrower than the succeeding part of the body ; 
but, when well extended, it is of equal breadth or very slightly broader. It is marked off 
by distinct though slight notches, and is indented in the middle line anteriorly ; I could 
not distinguish median and lateral papillse. The situation of the ganglia is marked by 
a reddish patch about one-sixth of an inch behind the anterior end. 

The broadening of the head would seem, therefore, taking all the descriptions 
into consideration, to be a variable characteristic, and hardly suitable for employment 
as a specific character, or for use as the diagnostic mark in a key. It will also be 



THE NEMEUTINES OF MILLPOUT AND ITS VICINITY. 25 

noted that, contrary to Burger, the Millport specimens showed a distinct marking 
off of the head from the body. 

The eyes appear to be another variable character. MacIntosh describes them as 
constituting a long wedge on each side, with the apex forwards. Joubin (8) speaks of 
" des amas de petits yeux noirs." I have, however, always found them arranged in 
a single line ; they are numerous, situated anteriorly on each side, along the junction 
of the pale margin of the head with the general brown of the dorsal surface ; they 
extend from near the tip to a point about half-way between this and the situation 
of the ganglia. 

The lateral grooves extend from nearly the anterior end of the head backwards 
to the level of the ganglia. They can be narrowed or closed, but the pit at their 
posterior end always remains open. 

Lineus ruber (Mull. ) ( = gesserensis auct. ). 

This is the commonest Nemertine of the shore at Millport ; it is found abundantly 
under stones at all tide-levels. I examined specimens from the shore near the Biological 
Station, from Balloch (l|- miles to the N.), from the islands in Millport harbour, and 
from Fairlie sands. 

Specimens live well in captivity in spite of lack of care in changing the water or 
in compensating evaporation ; they sometimes show a tendency to leave the dish 
and very often lie along the water-line at the side. A number of individuals commonly 
coil themselves up together. 

Specimens were met with of all lengths from 1 to 9 inches ; a common length is 
2 to 3 inches. The breadth does not vary so much as the length, being from 
1 to 2 mm. ; ordinary specimens thus appear moderately stout. 

The colour is remarkably variable. Joubin (8) has distinguished five colour 
varieties, as follows — (a) black and dark blue, (/3) dark olive green, (7) light olive 
green ; these three correspond to Lineus gesserensis as described by MacIntosh ; 
(<S) green and red, (e) red ; these two correspond to Lineus sanguineus in Macintosh's 
description. MacIntosh, for L. gesserensis, gives the two colours reddish brown and dull 
olive. Burger did not meet with this species at Naples. 

The specimens met with at Millport fall into three classes, as follows : — 

(i.) The commonest is a purple variety. This may be a pure deep purple through- 
out, though more usually the animal shows, when extended, a brownish tinge ; the 
colour may be mottled with a number of small patches of a lighter tint; and the 
ventral surface may be somewhat or even much lighter than the dorsal. 

(ii.) The next most common is a brown variety. The depth of tint varies, and a 
purplish tinge may be present, especially in the darker forms, due presumably to the 
cilia on the surface of the body. The head may be darker than the rest of the body. 
In a light-brown specimen the ventral surface was quite pale, and the lateral margins of 
the body were also marked by a narrow lighter stripe. 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 1). 4 



26 DR J. STEPHENSON ON 

As will be seen, these two varieties shade into each other, the purple having 
frequently a brownish, and the brown a purplish tinge. 

(iii.) The rarest variety is the green, of which specimens were obtained from an 
island in the harbour, and also from Fairlie sands. The colour may be a light green 
throughout ; or it may be darker, with a bluish tinge anteriorly ; or it may be a light 
olive, becoming reddish anteriorly (not due to the colour of the ganglia) ; or, finally, it 
may be a dark olive green, with a faint appearance of longitudinal striping. One 
specimen showed a white constriction at about the middle of its length ; and a number 
showed a pale lateral stripe along the margins of the body. 

Here, again, the mention of ' olive ' shows the tendency to a brown coloration ; but 
the green variety is, on the whole, much more sharply marked off from the others than 
the purple and brown from each other. 

In all varieties the margins of the head are pale. The numerous pale transverse 
wrinkles mentioned by Macintosh were as often absent as present. 

The head is somewhat flattened, and is marked off from the succeeding part of the 
body by slight notches. It narrows slightly, but only slightly, towards the anterior 
end, so that this extremity has a truncated appearance. In a few specimens a minute 
ciliated papilla was seen at the centre of the anterior end, and once a slight indentation 
was noticed at this spot ; usually neither was visible. The well-marked lateral grooves 
extend from the notches forwards to nearly the anterior end of the head. 

I have paid some attention to the comparative width of head and body in these 
forms, since this character is here, as in many other species of Nemertines (of. ant., on 
L. longissimus), commonly used as a diagnostic mark. I find that the head may be, at 
its widest part, either very slightly broader, or no broader, than the succeeding part of 
the body. Macintosh's expression on this point is that the head is " distinctly wider 
than the rest of the body," or " decidedly wider than the succeeding part of the body." 
Burger defines it as " ein wenig verbreitert." 

The eyes vary in number from, one to five on each side, or three to eight 
altogether. They are in the majority of cases situated in a regular longitudinal row at 
the upper edge of the pale lateral margin of the head, the anterior eye on each side 
being the largest. These two are symmetrically arranged, but the succeeding eyes are 
usually not symmetrical, nor the same in number on the two sides. 

The mouth is an elongated median slit, situated in a pale oval area, behind the level 
of the lateral notches. The proboscis reaches back about one-third of the animal's 
length. The intestine gives off small simple or bifurcated diverticula. 

Burger unites under L. ruber (Mull.) both L. gesserensis and L. sanguineus, 
described as separate species by MacIntosh. Joubin does the same, his first three 
colour varieties corresponding to L. gesserensis, his last two to L. sanguineus. 

Besides their colour, the two species gesserensis and sanguineus have been stated to 
differ in the comparative width of head and body, in the disposition of the eyes, and in 
their habits ; and it is interesting to note that, while agreeing in colour with gesserensis, 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 27 

the Millport specimens coincide rather with sanguineus in the fact that the head is 
scarcely wider than the rest of the body, in the usual arrangement of the eyes in regular 
rows, and also in the fact that they may coil themselves into a firm ball. The last 
character is, however, occasional rather than common ; the animals, when not aggregated 
together in a heap, usually lie along the water-line on the side of the dish. Thus, 
though I did not meet with any forms corresponding in colour to Macintosh's descrip- 
tion of L. sanguineus, the above facts would seem to confirm the propriety of uniting 
this with L. gesserensis under a common appellation. 

Micrura scotica, n. sp. 

One specimen only was found ; this was dredged in 1 5 fathoms off Ascog Bank. 

Length 1\ inches (about 60 mm.), breadth l l\ mm. ; body oval in transverse section, 
margins not flattened ; posterior end tapering ; in general appearance the specimen was 
not unlike Lineus ruber. 

In colour the dorsal surface was light brown with a purplish tinge, uniform over all 
but the posterior part, where the situation of the alimentary canal and its lateral 
branches appeared pigmented, the rest pale ; the margins of the body were white, and 
the ventral surface whitish ; there was a small red area near the tip of the snout, 
between the anterior eyes ; the area of the nerve ganglia was reddish, and the margins 
of the mouth were whiter than the rest of the ventral surface. 

The animal either coiled itself up into a ball when at rest, or lay in loose folds ; 
part of the body might be thrown into a spiral. In contracting itself, it contracted 
the anterior part most, so that this part of the body then appeared wrinkled. 

Locomotion was effected by a rapid gliding. The animal could not swim ; but 
frequently progressed easily and naturally in a backward direction, in its usual 
gliding manner and not by contractions of the body-wall. 

The head (fig. 17) was of the shape of an elongated triangle, with a blunt and 
rounded apex at the snout ; it was slightly marked off from the body, and was scarcely 
as broad as the succeeding region. The cephalic grooves extended along its sides in its 
whole extent ; the posterior part of the depth of the grooves was red, The anterior end 
of the head bore three retractile papillse, a dorsal or vertical, and a lateral on each 
side ; each was an elongated, ridge-like elevation, and the three were arranged so as to 
radiate from a common centre at the anterior end of the axis of the body. In the 
retracted condition of the head, a groove made its appearance on the anterior end of the 
ventral surface ; and this, conjoined with the retraction of the three papillae, gave the 
appearance of the crossing of a horizontal and a vertical groove at the tip of the snout. 

The eyes were arranged in two rows, one on each side anteriorly, near the lateral 
margin (fig. 17); they were difficult to see well and to count, as, except the anterior 
ones, they were small and were just under cover of the pigmented area, not in the 
marginal pale zone. There were eight on the left and five on the right side ; the 
anterior eye on each side was considerably larger than the others. 



28 DR J. STEPHENSON ON 

The mouth was a much elongated slit, situated behind the slight constriction which 
separated the head from the body. 

The tail (fig. 18) was a small, whitish, cylindrical submoniliform appendage of the 
posterior end of the body, '8 mm. long — hence visible to the naked eye without difficulty. 

Examined some months afterwards in spirit, the general colour of the specimen was 
a light brown, with no distinction between dorsal and ventral surfaces.' The cephalic 
grooves were very distinct ; the mouth appeared as a roundish hole. 

Examination of Sections. — The anterior end of the animal was sectioned in order 
to investigate certain other characters which have been used for diagnostic purposes in 
descriptions of species of this genus. Of these the following may be mentioned : — 

The cephalic grooves are deep ; they would have to be one-third to one-fourth deeper 
than is actually the case in order to reach the brain. The cerebral organ is at first on 
the outer side of the brain ; farther back it fuses with the dorsal ganglion, and then 
completely surrounds the hinder end of the latter as in a case. The dorsal ganglia are 
larger than the ventral. 

As usual in the genus, there is no diagonal muscular layer; and neurochord cells- 
are absent. 

I propose the following diagnosis for this form : — Length 65 mm., breadth 2|- mm. ; 
resembles Lineus gesserensis in appearance. Colour light brown, margins of head 
and body white, ventral surface whitish. Head slightly marked off from the body, 
not broader than the succeeding fart. Cephalic grooves along margins of head, 
reaching anteriorly to its tip. Eyes in tivo rows, one row anteriorly near margin 
on each side, 5-8 in each row, small and inconspicuous except the anterior ones. 



REFERENCES TO LITERATURE. 

(1) Benham, W. B., "The Platyhelmia, Mesozoa, and Nemertini " : Part IV. of A Treatise on Zoology, ed. 

Lankester, London, 1901. 

(2) Burgee, 0., " Untersuchungen liber die Anatomie mid Histologie der Nemertinen . . .," Zeitschrift 

filr wiss. Zool., 1., Heft 1, 1890. 

(3) Burger, 0., "Die Nemertinen des Golfes von Neapel," Fauna und Flora des Golfcs von Neapel, 

22 Monographic, Berlin, 1895. 

(4) Burger, 0., "Nemertini," Bronn's Klassen und Ordnungen des Tierreichs, Band iv., Supplement, 

Leipzig, 1897-9. 

(5) Burger, 0., "Nemertini," in Das Tierreich, Berlin, 1904. 

(Q) Coe, W. R., " Nemerteans of the W. and N. W. Coasts of America," Bull. Mus. Harvard, vol. 
xlvii., 1905. 

(7) Joubin, L., "liecherches sur les Turbellaries des cotes de France (Nemertes)," Arch. Zool. Exper. 

(2), viii., 1890. 

(8) Joubin, L., "Les Nt'mertiens," Faune Francaise, pub. par les soins de R. Blanchard et J. de Guerne, 

Paris, 1894. 

(9) Joubin', L., " N'mertiens," Blanchard's Trait/' de Zoologie, fasc. xi., Paris, 1897. 

O.0) MacI.ntosii, W. C, A Monograph of the British Annelids, Part I.: "The Nemerteans," London,. 
1873-4. 



■i-' 



Trans. Roy. Soc. Edin' 



Stephenson : The Nemertines of Millport. 



Vol. XLVIII. 



n 



A*x > 




■ COMK. 

- m 



h 







Fis .£> 



%4 




f'^-l 




FCtj.if 



Fig. * 




f^.ii Fi 3 f i 






**-... 





Fi*. /8 



J.S.*i. 



^ • /3 



% 7 



Ow Watsrstons Scroa.Lith, Edinburgh.. 



THE NEMERTINES OF MILLPORT AND ITS VICINITY. 29 

(11) Oudemans, A. C, "The Circulatory and Nephridial Apparatus of the Nemertea," Quart. Journ. 

Microsc. Science, vol. xxv., Suppl., 1885. 

(12) Oxner, M., "Sur quelques nouvelles especes des Nemertes de Roscoff," Arch. Zool. Exper. (4), vi. 

(Notes et Revue), 1907. 

(13) Punnett, R. C, "Addenda and Corrigenda to the Nemertini," in Part IV. of A Treatise on Zoology, 

ed. Lankester, London, 1901. 



EXPLANATION OF FIGURES. 
(Figs. 3, 4, 5 drawn from sections by Zeiss's Abbe's drawing apparatus.) 

Fig. 1. Head of Tubulanus annulatus from the dorsal surface, to show appearance of grooves and cerebral 
organs ; pigmented areas indicated. 

c, cerebral organ; /., frontal pigmented area (Stirnfeld) ; g., cephalic groove. 
Fig. 2. Part of the left lateral margin of the head of another specimen of the above ; the anterior wall 
of the cephalic groove, and the circular depressed area (cerebral organ) are lighter in tint than the surround- 
ing parts. 

c, cerebral organ; g., cephalic groove. 
Fig. 3. Transverse section of head of Tubulanus annulatus, through the left cerebral organ. x 55. 

c, cerebral organ ; br., brain. 
Fig. 4. Longitudinal section of head of Tubulanus annulatus. x 64. 

comm., commissure between the two halves of the brain; gl., head-gland; m., mouth; pr., pro- 
boscis ; pr. cav., rhynchoccelom ; x., cushion of cells described in text ; y., the smaller 
ventral cushion. 
Fig. 5. The dorsal cushion of cells in the proboscis of Tubulanus annulatus, more highly magnified. 
x225. 

Fig. 6. Stylet of Emplectonema gracile. 

Fig. 7. Usual form of stylet of Ampliiporus lactifloreus. 

Fig. 8. Exceptional form of stylet of Ampliiporus lactifloreus. 

Fig. 9. Dorsal surface of head of Ampliiporus pulcher. 

a.d.g., anterior cephalic groove, dorsal; p.d.g., posterior cephalic groove, dorsal; set'., secondary 
groove. 
Fig. 10. Ventral surface of head of Ampliiporus pulcher. 

a.v.g., anterior cephalic groove, ventral ; d., dimple in the course of the above ; p.v.g., posterior 
cephalic groove, ventral; sec, secondary groove. 
Fig. 11. Stylet of Ampliiporus pulcher. 
Fig. 12. Ampliiporus elongatus, to show the general form. x If. 

a., anterior end; p., posterior end. 
Fig. 13. Head of Ampliiporus elongatus, as a transparent object. 

a.e., anterior eyes; a.g., anterior cephalic groove; br., brain; c, cerebral organ; p.e., posterior 
eyes ; p.g., posterior cephalic groove. 
Fig. 14. Stylet of Ampliiporus elongatus. 
Fig. 15. Stylet of Prostoma candidum. 
Fig. 16. Stylet of Oerstedia dorsalis. 

Fig. 17. Head of Micrura scotica, from the dorsal surface; the position of the mouth on the ventral 
surface is, however, also indicated. 

e.g., extent of cephalic groove; e., eyes; g., reddish area due to brain; m., relative position of 
mouth ; x., line indicating junction of the pale margin with the pigmented area of the head. 
Fig. 18. Posterior end of Micrura scotica, to show relative size of tail. 



TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 1). 



( 31 ) 



II. — On some littoral Oligochseta of the Clyde. By J. Stephenson, M.B., D.Sc. 
(Lond.) ; Major, I.M.S. ; Professor of Biology in the Government College, Lahore. 
Communicated by Professor D'Arcy W. Thompson, C.B. (With Two Plates.) 

(MS. received August 11, 1910. Read November 21, 1910. Issued separately May 16, 1911.) 



CONTENTS. 



Introduction .... 
Tubifex costatus, Clap. 
Marionina semifusca, Clap, . 
Lumbricillus subterraneus, Vejd. 
Luinbricillus tuba, n. sp. 
Lumbricillus viridis, n. sp. 
Enchytraeus nodosus, n. sp. . 



PAGE 

31 
33 

35 
39 
42 
46 
50 



Enchytraeus dubius, n. sp. 
Enchytraeus sabulosus, Southern 
Enchytraeus albidus, Henle 
Fridericia bulbosa, Rosa 
Literature .... 
Explanation of Plates . 



PAGE 

54 
58 
59 
61 
63 
63 



Introduction. 

The following paper contains an account of certain of the littoral Oligochseta of the 
Firth of Clyde, found at and near Millport on the Island of Cumbrae, and at Wemyss 
Bay on the mainland. The investigation was begun during a two months' stay at 
Millport, from May to July, 1909, at which time I was working in the Marine Biological 
Station there, and it was completed after my return to India. I have to thank 
Mr R. Blmhirst, the Superintendent of the Station, for his constant kindness, and for 
the courtesy with which he placed all the appliances and resources of the Station at 
my disposal. 

In 1906, according to Southern (12), of the more than one hundred species of Enchy- 
trseids then known, only twelve had been recorded from Great Britain, and only seven 
from Ireland. The number of known species of Enchytrseids, as well as of other 
limicolous Oligochseta, has been considerably increased since then, and, mainly owing 
to Southern (12, 13, 14), this portion of the fauna of the British Isles has also become 
better known ; but, as that author remarks, " the large number of new species and of 
additions to the British list shows how much work remains to be done on this order 
before our knowledge can be considered in any way complete." 

The ten species of which an account follows are, with the exception of one Tubificid 
[Tubifex costatus, Clap.), all Enchytrseids. I need not say that the list is very far 
indeed from exhausting the littoral Oligochsete fauna of the Clyde ; it contains, indeed, 
only those forms to which I have devoted a fair amount of attention. The several 
species are, I think, interesting in various ways — some {Lumbricillus tuba, L. viridis, 
Enchytraeus nodosus, E. dubius) because they are new ; others [Tubifex costatus, 
Marionina semifusca, Lumbricillus subterraneus, Enchytraeus sabulosus) because, 

TRANS. ROY. SOC. ED1N., VOL. XLVIII. PART I. (NO. 2). 6 



32 DR J. STEPHENSON ON 

though not new, they have hitherto been recorded but once or a few times, and there 
are still lacunae to be filled before our knowledge can be considered complete ; or lastly, 
as Enchytrseus albidus, a well-known and widely distributed form, but at the same 
time one which appears to be very variable, and the Clyde specimens of which 
show peculiarities of their own. 

One further point may perhaps be noted here, rather than in the body of the paper 
— namely, the relationship between the genera Lumbricillus and Enchytrseus. 

The characters serving to distinguish these genera may be said to be four, (i.) The 
setae of Lumbricillus have a double, or J-shaped curve ; those of Enchytrseus are straight, 
with, however, a small hook-like curve at their proximal end. (ii.) The presence or 
absence of a penial bulb ; Eisen (4) not only makes this a distinguishing feature between 
the genera, but proposes it as a chief means of dividing the Enchytraeidae into two sub- 
families, the Lumbricillinae which have, and the Enchytraeinae which have not this 
structure (cf. below, under Enchytrseus nodosus and Lumbricillus viridis). (iii.) The 
' copulatory glands ' or ' Bauchmarkdriisen,' are supposed to be distinctive of the genus 
Lumbricillus; thus Beddard (1) writes: "One of the most characteristic structural 
features of the genus, though confined to a few species, is the outgrowths of the ventral 
nerve-chord in certain segments." (iv.) The multilobed testes of Lumbricillus are also 
one of its chief generic characters. 

If certain of the forms here described be examined with regard to these characters, 
the following conditions are found : Lumbricillus viridis, while in the other points 
showing a typical Lumbricilline structure, has, in the anterior part of its body, setae of 
the typical Enchytrseus form, while the posterior setae show only a very faint double 
curvature. Enchytrseus nodosus, though its setae are for the most part typically those 
of the genus in which I have placed it, shows in certain cases setae with an indication 
of a double curvature, copulatory glands, and a penial bulb, i.e. Lumbricilline characters 
are present ; the testes, however, are single on each side. Enchytrseus dubius has setae 
which are throughout similar to those described as typical for the genus to which I have 
referred it ; while in the possession of lobed testes, copulatory glands, and a penial bulb 
(though this latter is bifid internally), it agrees with Lumbricillus ; I may add that it 
has red blood, a feature commoner in the species of Lumbricillus than in those of 
Enchytrseus. Finally, Enchytrseus albidus, a very fairly typical species of its genus, 
has nevertheless an imperfect penial bulb, surrounded, it is true, by other and smaller 
aggregations of gland cells (cf. below, under E. nodosus). 

It would therefore appear that not only are the two genera closely allied, but that a 
number of intermediate forms exist which serve to bridge over the interval between 
the two.'"' 

* With regard to the existence of copulatory glands in the genus JUnchytrceus, Southern has recorded their 
occurrence in K. Idbai/us ; and there is also a penial bulb in this species (14). 



SOME LITTORAL OLIGOCH^TA OF THE CLYDE. 33 

Tubifex costatus (Clap.). 

This worm was first very briefly described by Claparede in 1863. It was rediscovered 
by Be nh am in material from Sheerness in 1891, and was fully described by him (2), 
especially with regard to its setse and genital organs, in a paper with many excellent 
illustrations. 

The species was placed in a separate genus, Heterochwta, by its original discoverer, 
as well as by Benham ; and this distinction is also assigned to it by Vejdovsky (16) and 
Beddard (1). Michaelsen, in the body of his work on the Oligochseta (11), includes 
it in the genus Psammoryctes, but in the appendix (p. 522) unites this genus with 
Tubifex, and the worm thus becomes Tubifex costatus. 

Southern (14) records it from between tide-marks on the Irish coast, but gives no 
description ; he refers to its mention by Friend, in a paper which I have not seen 
(Irish Nat., 1897). Evans (5) records it from the Haddingtonshire coast. 

The worm is thus, apparently, described with any degree of completeness only in 
Benham's paper ; a few additional particulars, and an account of one or two features in 
which my specimens differ from Benham's, may therefore be of interest. 

The worms were found at Fintry Bay, about high- water mark, under moist stones, 
at a place where fresh water was running down on to the shore. They live for days 
in half salt, half fresh water. Their average length was greater than that of pre- 
vious records, being about an inch (Claparede 16 mm., Benham f of an inch) ; 
specimens were met with up to an inch and a quarter. In colour they were of various 
shades of red, the anterior part of the body being paler ; as also the genital segments 
on account of the presence of genital products. The number of segments was sixty- 
three to sixty-seven (about forty, Benham). 

The detailed account of the setse given by Benham must exhaust the subject. 
Briefly, the setse are all of the ordinary double-pronged type, except those of certain 
dorsal bundles in the anterior part of the body, where they are ' palmate ' (segments 
v.-xiv.). I may add that in length the ventral setse in the anterior part of the body are 
about "11 mm., in the posterior about '086 mm. ; the palmate setse average '095 mm. 
The numbers per bundle in the Millport specimens were rather greater than those found 
by Benham ; thus in the ventral bundles there were up to seven in the anterior part of 
the body, not more than two or three in segments x.-xv., and posterior to this four, three, 
two, or one only ; the palmate dorsal setae were in bundles of six to thirteen, the double- 
pronged dorsal setse posterior to these in bundles of four, three, two, or one, like the 
corresponding ventral setse. 

Benham looked for a long time in vain for intermediate forms between the two types 
of setse. These ' multidentate ' forms are very common in the Millport specimens ; some 
are figured in fig. 1. As to their distribution, they are found in the segments in front 
of and behind those containing the palmate setse ; thus the most anterior dorsal bundles 
(ii. and iii.) may either consist of the usual doubled-pronged setse or of these irregular 



34 DK J. STEPHENSON ON 

forms. In segments xiv.-xviii. doubled- pronged and palmate setae may be mixed ; or the 
bundle on one side may consist of double-pronged, on the other side of palmate setae. 

The circulatory system, briefly referred to by Benham, deserves description (PI. I. 
fig. 1). The dorsal vessel is connected with the alimentary wall and covered by 
chloragogen cells as far forwards as the tenth segment, where it becomes free ; from the 
seventh to the second segment it gives off prominent lateral loops, non-contractile, 
tortuous, running on the inner face of the body-wall. The dorsal vessel bifurcates at 
the junction of prostomium and first segment ; the branches unite again below about the 
level of the setae of segment iv. The ventrally situated vessel which is thus formed is 
outside the chloragogen cells ; it unites posteriorly with the subintestinal in the eighth 
segment. 

The supraintestinal vessel is present on the alimentary canal, covered by chloragogen 
cells, from about the place where the dorsal vessel leaves the intestine to the fifth 
segment anteriorly. It gives origin in segment viii. to the two hearts, greatly dilated 
vessels, one on each side, which contract from above downwards, and, as Benham has 





Fig. 1. — a, a multidentate seta from the fourth segment of a specimen of Tubifex costaius, 
b, intermediate forms of setse from the fourth segment of a specimen of Tubifex 
costatus . 

remarked, alternately ; these hearts, approaching each other ventrally at the level of 
septum ■§, are prolonged almost parallel to each other, without immediately uniting, 
backwards through the ninth segment; they then join to form the ventral vessel, 
which is continued backwards below the alimentary tube in the body cavity. 

There remains to be mentioned the subintestinal vessel, a single median channel, on 
the intestinal wall within the investment of chloragogen cells. This can be distinguished 
in sections as far back as segment xiv. ; in segment xii. it is equal to the ventral vessel 
in size, in segment x. larger ; it soon becomes small again, and dies away on the intestine 
in the anterior part of the eighth segment, after receiving the posterior end of the 
ventrally situated vessel previously described. The relations of these several vessels are 
illustrated diagrammatically in PI. I. fig. 1. 

The parietal plexus is most copious in the posterior part of the body ; the loops 
branch and reunite on the inner surface of the body-wall, but do not penetrate the 
circular muscular coat ; these branches on the body-wall are of considerable size — 
indeed, are of the full diameter of the loop which gives origin to them before it 
divides up. 

The shape of the cerebral ganglion is sometimes made use of in specific diagnoses. 



SOME LITTORAL OLIGOCH^ETA OF THE CLYDE. 35 

It may be mentioned that it is about as long as broad, is somewhat narrower behind 
than in front, and is slightly indented posteriorly. It is contained in segment i. 

With regard to the genital organs, only certain points in connection with the 
atrium and spermatheca need be considered. The atrium, according to Benham, shows 
a division into two parts, which he distinguishes as glandular and non-glandular, of almost 
equal extent, the lining cells of the first part being cubical and vacuolated, as if a 
secretion had been discharged, while the cells of the second part are flat. In my 
specimens also, two regions are to be distinguished ; but the first region is very much 
less extensive than described by Benham, and extends only for a very short distance 
on both sides of the entrance of the prostate ; the prostate enters the atrium almost 
immediately beyond the ending of the vas deferens (cf. Benham's fig. 18), and the 
glandular cells extend about equally on both sides of this point. In character these 
cells are tall and filled with deeply staining granules, but not vacuolated. 

The sperm athecse present an external portion, narrowing gradually towards the 
aperture, with a vertical position in the segment, and a long, more dilated, sausage- 
shaped cavity, bent into a number of curves ; the whole being either confined to 
segment x., or extending forwards into ix., or backwards to the level of xii. This second 
internal and far more extensive portion is, in my specimens, lined by tall columnar cells 
of large size, extensively vacuolated ; the vertical portion, or duct, in extent about half 
the vertical diameter of the segment, is lined by more solid-looking smaller columnar 
cells, the outlines of the individual cells being often indistinguishable. 

Marionina semifusca (Clap.). 

This worm was first described in 1861 by Claparede (3), who discovered it in the 
Hebrides. His account deals almost entirely with the reproductive organs ; beyond 
this it includes only a few short statements as to size, colour, nephridia, and ccelomic 
corpuscles. Southern (13, 14) has recently recorded the same species in both Ireland 
(Dublin Bay) and Scotland (Dalmeny, where the specimens were collected by Evans), 
and has given (14) further particulars of its anatomy. The following account deals 
principally with points which have not yet received detailed attention. 

The worms were found at Fintry Bay, about high- water mark, under moist stones, 
at a place where fresh water was running to the shore ; and subsequently at Balloch, in 
a similar locality. 

Length 16 mm. Segments forty -two. Colour light red, whiter in front of the 
clitellum. Both ends blunt ; head-pore at the junction of prostomium and segment i. ; 
clitellum embracing segments xii. and xiii. 

The setse are of the same character throughout, in both dorsal and ventral bundles. 
They are slightly curved in a {-shape, the distal curve, however, being much less in 
extent than the proximal, which is a long, gentle sweep ; they are comparatively slender, 
and pointed at; both ends (fig. 2). In number they are, in front of the clitellum, 



36 



DR J. STEPHENSON ON 



usually six, varying from five to eight ventrally, and four, five, or six dorsally ; in the 
post-clitellial segments the numbers are about the same, except that nine were once met 
with in one of the ventral, and seven in one of the dorsal bundles. Ventral setae are 
absent in segment xii., in which also the dorsal setae are few or absent. 

The length of the setae is on the average about "1 mm. ("095 — "108 mm.). There 
is no appreciable difference in length between ventral and lateral setae, or between 
those in the anterior and those in the posterior part of the body. There is, however, 
a difference between the various setae of a bundle ; the setae are disposed fan-wise in 
each bundle, and the outer setae are rather longer than the inner, the length de- 
creasing from the outer to the inner side with some regularity. In illustration, 




Fig. 2. — A setal bundle of Marionina semifusca. 



the following figures may be given (cf. also fig. 4) ; the numbers represent the 
lengths of successive setae from outer to inner side of the bundles : — 



Ventral setje 



Dorsal setsei 



(•099 
099 
101 
101 

099 

0945 
0967 
101 



•099 
•099 
•101 
■101 

•0945 
•0922 
•0967 
•0967 



09 
0967 
101 
099 

0855 
09 

0922 
09 



0855 


•0855 


•0855 


09 


•09 


•09 


101 


•099 


•099 


099 


•0945 




09 






0922 






09 






0877 


•0832 





As regards the alimentary canal and its appendages, the pharynx has the usual 
form, and occupies segments ii.-iii. ; the oesophagus is narrow, and begins to be clothed 
by chloragogen cells in segment iv. ; the tube, though still narrow, dilates a little 
in segments ix.-x. ; it is again very narrow in the genital segments, and finally swells 
and assumes the usual characters of the intestine in segment xiv. There are no pepto- 
nephridia. As remarked by Southern, the septal glands are large (PI. I. fig. 2) ; they 
extend farther back than usual, one pair being situated always in segment vii. 
(cf. Southern), and there may be a pair in segment viii. The chloragogen cells are 
of a very decided brown. 

The dorsal vessel begins in segment xiv. (xiii. Southern), and bifurcates at the 
junction of prostomium and first segment ; the two branches into which it divides reunite 
ventrally in segment iv. to form the ventral vessel. The lateral commissural vessels 
are four on each side ; the first begins above in segment ii. and ends below in the 



SOME LITTORAL OLIGOCH^ETA OF THE CLYDE. 37 

anterior part of iv. ; the second begins above, near the junction of iii. and iv., and ends 
below, just behind the first ; the third is wholly contained in segment iv., the fourth in 
v. The last two join the ventral vessel below, the first two join the branches which 
unite to form the ventral vessel. 

The ccelomic corpuscles are round or broadly oval, disc-shaped, and granular ; they 
are of large size, measuring in the fresh state from 22 to 36 m ; as seen in sections, 
however, they are smaller, and average 20 m, the largest measured being 25 m. They 
are not obviously nucleated in the fresh condition ; the nucleus is conspicuous in stained 
preparations, lying in the middle of a loose reticulum. They are very numerous, and 
the body-cavity may be crowded with them. 

The nephridia begin in segment v., but are absent in xii. and xiii. The ante-septal 
portion consists of the funnel only ; the post-septal is a large ovoid mass, coloured by a 




Fig. 3. — Cerebral ganglion of Marionina semifusca. 

brown pigment in its anterior half; the tube is loosely coiled within the mass of the 
organ. The external openings are in this species easily visible when a specimen is 
examined from the ventral surface ; they are in front of the ventral setse. According 
to the evidence of sections, the duct comes off from the mass of the gland well in front 
of the middle, though this was not made out in the living specimens ; but both modes 
of examination show that it runs backwards to the external aperture, not forwards, as 
figured by Claparede for his specimens. 

The cerebral ganglion is one-and-a-half times as long as broad ; anteriorly the 
margin is straight, posteriorly the ganglion is deeply indented (fig. 3). The ganglia 
on the ventral nerve-cord are conspicuous, especially in the anterior part of the body. 
" Copulatory glands " occur in the neighbourhood of the genital segments (cf. Southern). 

The testes are large, equal in length to the anterior half of segment xi. ; they are 
somewhat triangular in shape, attached by their narrow end, with the base of the 
triangle directed posteriorly. The funnel is short, ovoid, one-and-a-half times as long as 
broad, with an obvious lumen in which ciliary action is very visible ; in general structure it 
resembles related forms. The sperm morulas appear to be confined within the limits of 
segment xi., septum -j-J being slightly bulged backwards. The vas deferens is thin, 



38 DR J. STEPHENSON ON 

much coiled, and may extend backwards behind the clitellum. The penial gland in my 
specimens seems to differ from the descriptions of Claparede and Southern. Accord- 
ing to the former, " ils occupent la cavite periviscerale en entier, dans le onzieme 
segment, produisant meme souvent une dilatation du corps dans cette region " ; in the 
diagnosis of the species they are called " enormes," and in the figure are shown as being 
kidney-shaped. Southern calls them 'large,' and describes and figures them as 
cylindrical. 

I find that they are somewhat flattened ventrally where they are sessile on the 
bodv-wall, but for the rest are spherical, and my specimens seem to give no hint of 
either a kidney-shaped or cylindrical form. They are large, as in most related species, but 
not, I think, so large as to call for any special remark ; and in my specimens they do not 
by any means fill the coelom in their segment, nor cause a bulging of the body-wall. 

The ovaries are smaller than the testes, and are attached as usual to the posterior 
face of septum \^. The ova, when detached, are seen in sections free in segment xii., 
and are confined to this ; in the living they appear at the level of segment xiii. , perhaps 
through the bulging backwards of the septum. The funnel is small, and the oviduct 
short. 

The spermatliecze consist of an ampulla, duct, and gland-cells, having the ap- 
pearances and relations described by Claparede. The ampulla is continuous with the 
oesophagus, but I have not been able to trace a continuity of lumen between the two. 
The wall of the ampulla is thin, and the cells composing it are flattened. The cells of 
the wall of the duct are in a single layer, and are not covered externally by muscle-fibres. 
The gland-cells at the base of the duct are continuous on the one hand with the epithelial 
cells of the surface of the body, and on the other with those of the duct ; the muscular 
layer of the body-wall is continued amongst and between them, the cells being so much 
elongated that they extend inwards a considerable distance beyond the level of this 
muscular layer. 

The clitellum appears as a mixture of clear and hyaline areas. Clitellar cells are 
absent over the situation of the penial glands. 

Sporozoa occur in the oesophagus. 

As Beddard (1) remarks, there would seem to be a mistake in Claparede's descrip- 
tion of the gonads ; the testis he places in segment x., and the ovary in xii. (i.e. xi. and 
xiii. according to our notation). I do not understand, also, how he comes to speak of 
both testis and ovary as being single ; he is evidently speaking of the glands them- 
selves, not of the aggregations of sperm morulse or ova ("les organes sont fixes par 
un pedoncule a la paroi du corps [or, rather, to the posterior face of the respective 
septa] ... les produits, savoir les zoospermes et les ceufs, tombent, une fois arrives a 
maturite, dans la cavite periviscerale"). 

1 do not consider the identification of the above form with Claparede's species to be 
absolutely certain, since Claparede's description is incomplete, and, in regard to the 
points mentioned by him, there are a number of differences to be taken into account. 



SOME LITTORAL OLIGOCH^TA OF THE CLYDE. 39 

The length is perhaps not very important, but the direction of the duct of the nephridia, 
and the shape and size of the penial glands may also be mentioned. It is principally the 
characters of the spermathecal apparatus which have determined me to identify my 
specimens with his description. 

Lumbricillus subterraneus (Vejd.). 

Under the name Pachydrilus subterraneus, Vejdovsky (15), in 1889, described an 
Enchytraeid which he had first found in a well at Prague ; some worms sent to him 
subsequently from Lille, where they had been discovered in the water-pipes of the town, 
were found to belong to the same species. The next record of the occurrence of this form 
is by Southern (14) ; a large number were sent to him from the sewage works at 
Belfast, and the same species was also found by him in a stream in Lancashire which 
was excessively contaminated with trade effluents. In the present case, the third record, 
the worms occurred on the seashore, about high- water mark, where they must at times 
be exposed to the influence of salt water. 

Though there can be no doubt about the identity of the present form with that 
described by Vejdovsky, I add here a number of anatomical particulars, since in certain 
points the original description is somewhat brief. This is the case, for example, with 
regard to the setae ; the structures known as " copulatory glands," or " Bauchmarkdriisen," 
were also not described by Vejdovsky (and may therefore have been absent) ; in con- 
sequence, this species is represented by Beddard (1) (p. 325, in the key to the various 
species of the genus) as not possessing them, which might possibly lead to some confusion ; 
they were present in Southern's specimens. 

Found about high- water mark, Fintry Bay, under moist stones, at a part where fresh 
water was running to the shore ; and again at Balloch. The animals live well for 
several days in a mixture of equal parts of salt and fresh water, and equally well in 
altogether fresh water. 

Length up to 1 inch (25 mm.) ; fairly stout, tapering towards both ends, most 
gradually towards the anterior end. Colour, various shades of red, whiter about the 
genital region ; ova visible as distinct pinkish- white masses. Locomotion by wriggling. 

Segments forty-nine to fifty- seven. Prostomium blunt, with a number of small 
papilliform projections. Clitellum includes segments xii.-xiii., and may encroach on xi. 
Setse. — The dorsal series are dorsal, not lateral, in position ; in the ante-clitellial 
segments they number four to eight per bundle ; in segment xii. fewer, two or three ; 
in the post-clitellial segments three to seven. The ventral series number five to eleven 
in the ante-clitellial, three to eight in the post-clitellial segments ; there are no ventral 
setae in segment xii. The setae are of the same type in the two series of bundles ; each 
seta (fig. 4.) is J-shaped, the proximal of the two curves being the more gradual ; each 
is moderately stout, and is thickest about the middle of its length, but there is no 
distinct nodulus ; the point is single. The setae of a bundle are arranged fan-wise, and 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 2). 7 



40 



DK J. STEPHENSON ON 



the outer setae of a bundle are longer than the inner, each bundle forming a gradated 
series ; thus the lengths of the setae of a bundle of five in segment x. were, from the 
outer to the inner side, 106, 101, 97, 90, 81m; in another bundle the lengths were 104, 




Fig. 4. — A setal bundle of Lumbricillus subterraneus. 



100, 94, 90, 83 /«. The longest setae are thus rather more than "1 mm. in length; 
in thickness they are about 5-6 m. The ante-clitellial of both dorsal and ventral series 
are on the average rather longer than the post-clitellial. 

The pharynx occupies segments ii.-iii. The oesophagus begins to be covered with 
chloragogen cells in vi. ; it presents no dilatations or diverticula, and passes into the 
intestine about segment xvi. ; this latter portion of the alimentary canal is bulged 
interseptally and constricted at the septa. Septal glands (PL I. fig. 3) are present in 
segments iv., v., and vi. ; those of each side are, as usual, longitudinally connected ; the 
glands in iv. are small, in v. are spread out on the septum (-|), and in vi. are large, 
lying mainly longitudinally in the segment. 

The dorsal vessel varies somewhat in its place of origin ; this may be from the 
thirteenth to the seventeenth segment ; it bifurcates in the prostomium. There are 
four lateral commissures on each side ; the first of these arises dorsally in the posterior 
part of segment iii. and runs forwards into ii., or even i. ; the second arises dorsally in 
the anterior part of iv., and runs forwards into iii. ; the third belongs altogether to 
iv., and the fourth to v. The ventral vessel bifurcates anteriorly in iv. ; it exists as a 
definite vessel as far back as the anus, being frequently separated from the intestine, so 
that a "window" intervenes between the two. The appearance of the 'sinus 1 in the 
alimentary wall when the posterior part of the body was much engorged, the animal 
dying and the blood coagulating, was that of a thick, close-set network of large vessels, 
with hardly any interspaces between them. The blood is red. 

The nephridia begin in segment vii. ; each is an ovoid mass, with a very small 
ante-septal portion ; there is a slight brown pigmentation in the anterior part of the 
post-septal. The tube is loosely coiled within the mass ; the duct leads forwards from 
its lower surface in front of its posterior end, and is much shorter than the body of the 
nephridium. The codomic corpuscles are irregularly pear-shaped or oval granular 
nucleated cells. There are also to be seen in the body-cavity a number of smaller 
spherical refractile bodies. 



SOME LITTORAL OLIGOCH^ETA OF THE CLYDE. 41 

The cerebral ganglion is rather longer than broad, indented posteriorly but not 
anteriorly ; two small dark dots are to be seen, one on each side, in its posterior part 
(cf. Enchytrseus albidus). Ganglionic swellings are well marked in the anterior portion 
of the ventral cord, but some distance behind the genital segments they become 
scarcely noticeable, the nerve cells being distributed over the whole length of the cord. 
The cord encloses a " central canal." 

The " copulatory glands " are present in segments xiii. and xiv. as conspicuous lobed 
masses around the ventral cord, their centre at a level just posterior to the insertion of 
the setae (PI. I. fig. 4). Each gland consists of a mass of large cells, spherical, pear- 
shaped, or polygonal, which is situated ventral to and on each side of the cord. These 
cells do not quite meet above the cord, but, connecting those of one side with those 
of the other over the dorsal surface of the cord, there appears in sections a deeply 
staining band. The dorsal surface of the cord is indented between the cells of the two 
sides in the middle line. 

The glands are attached to the ventral body -wall by an almost homogeneous stalk ; 
the circular muscular coat appears interrupted at intervals along the area of attachment 
(the fibres being presumably displaced), so that the stalk of the gland seems here to 
fuse with the epidermis. The body of a number of these cells stains very lightly, and 
shows an appearance something like that of an empty reticulum ; the nuclei of the 
gland-cells stain evenly ; those of the nerve ganglion cells which immediately surround 
the cord show, on the contrary, a number of distinct granules of chromatic material. 

The testes form a bunch of pear-shaped masses, attached by their narrow ends to 
septum TT ; they shift with the movements of the worm, appearing now on one side, 
now on the other side of the septum. Their products get forward into segment viii., 
and fill segments ix. and x. 

The sperm funnels may be as much as nine times as long as broad. They are, 
however, here as in other species very contractile, and may shorten (e.g. on teasing) 
to as little as twice as long as broad. Even in the body they may appear only about 
four times as long as broad, and vary. The vas deferens does not extend posteriorly 
beyond the clitellum ; it forms a fairly small coil, equal when uncoiled to about half 
a dozen segments ; it has a fine lumen, in which, in teased specimens, active ciliation 
can be seen to be going on. The male apertures appear as semicircular fissures, 
convex towards the middle line. The penial bulb is a spherical mass of considerable 
size, its diameter about a quarter of the whole diameter of the body ; the vas deferens 
penetrates the bulb laterally to its centre ; the bulb is attached to the body-wall by 
a strand of tissue which passes dorso-laterally upwards from the upper surface of 
the mass. 

The ovaries are on the posterior face of septum \^ ; ova are found as far forwards 
as segment viii., and backwards far behind the clitellum. 

The spermathecal apparatus in the living animal consists of a somewhat spindle- 
shaped mass, in which ampulla and duct are not to be distinguished ; a mass of 



42 DR J. STEPHENSON ON 

glandular cells surrounds the aperture. On examining a series of sections, the ampulla 
is found to comprise the internal half of the mass ; it is ovoid or somewhat pear-shaped, 
communicating with the oesophagus by its narrow end ; its walls are lined by a low 
cubical epithelium, and in the lumen, arranged as a layer all round, are usually numerous 
deeply staining heads of spermatozoa. The duct, or outer half of the mass, is not 
sharply delimited from the ampulla, though often, in sections, appearing to be separated 
from it by a kink in one or other wall ; its lumen is narrow, and it is lined by high 
columnar cells ; it has a well-marked muscular investment. The gland-cells near 
the external aperture are really the lining cells of the duct, which here extend outwards, 
breaking through the muscular investment of the duct, which can still be seen in 
places between the cells ; their nuclei are peripherally situated, outside the muscular 
layer, and the cells are continuous at the orifice with the surface epithelium. The 
above details are shown in PI. I. fig. 5. 

Certain parasites (Gregarines) are seen in the body-cavity in several specimens. The 
body of these sporozoa is dark and opaque, their nucleus clear ; the length of the 
double animal is about '5 mm. 

The alimentary canal also usually contains numerous sporozoa, with a much- 
elongated, deeply staining nucleus ; the whole width of the lumen of the alimentary 
tube may be packed with them. In one series of sections one of these forms is present 
in the spermatheca, into which it had probably wandered from the oesophagus. 

The specimens of this species sent to Vejdovsky from Lille differed from those 
previously obtained in Prague in having gland-cells round the apertures of the 
spermathecse. It will be seen that in this respect the Millport specimens agree with 
those from Lille ; Southern's specimens also possessed these glands. It is noteworthy 
that the same or a similar parasite should occur both in Vejdovsky's specimens 
and mine ; in mine, however, two individuals were commonly found joined together, 
which appears not to have been the case in the previous specimens. 

Lumbricillus tuba, n. sp. 

Common ; found about high-water mark, Millport. 

In length this species is from -|" to 1^" ; it is tapering at both ends, more so anteriorly. 
Its colour is pale pink, the anterior half lighter than the posterior ; ova may be seen 
as brilliant white spots ; the whole animal is fairly transparent. The worms move 
when disturbed in an active, wriggling, nematoid manner. 

The number of segments varies within only narrow limits — thirty-five to thirty-nine. 

I'rostomium blunt ; head-pore present, but no dorsal pores. 

The setae are of the same type in both ventral and lateral series ; they are somewhat 
{-shaped, but the distal curve is very slight ; they are comparatively slender ; there is 
no definite nodulus, but the shaft is slightly thicker a little distal to its middle. They 
are arranged in a fan-like manner in each bundle ; the outer setae of a bundle are not, 
as in L. subterraneus, longer than the inner. 



SOME LITTORAL OLIGOCH^TA OF THE CLYDE. 43 

The ventral setae are usually four to six (occasionally seven) in a bundle — commonly 
six in the anterior, five in the posterior part of the body ; there are no ventral setae 
in segment xii. The lateral setae are three to five in a bundle, except that in segment 
xii. there are only two, or one, or none. 

In length the setae are about '07 -'08 mm., the ventral being on the whole a little 
longer than the lateral. 

The alimentary canal has the usual relations. Septal glands are present in con- 
nection with septa f , f , and f- ; they are enclosed within the septa, which split to contain 
them, and thus suspend them to the body-wall ; they are less bulky than in some other 
species. Chloragogen cells begin in segment vi. ; they are very finely granular. There 
are no peptonephridia. The ossopJiagus, narrow as far as segment vii., dilates in a 
fusiform manner from vii. to x., and in this region it may be intersegmentally constricted 
like the intestine ; it is narrow in the genital segments, and widens to form the intestine 
in xv. 

The dorsal vessel begins in the thirteenth, fourteenth, or fifteenth segment, and 
bifurcates at the junction of the prostomium and first segment. The ventral vessel is 
distinct throughout the body, and bifurcates in segment iv. There are four pairs of 
lateral commissures ; the first originates from the dorsal vessel in segment iii. and 
passes forwards into segment ii. ; the last belongs to the fifth segment. The blood is a 
light yellowish red. 

There are numerous body -cavity corpuscles, nucleated and granular, mostly of the 
form of circular, bean-shaped, or elongated pear-shaped discs ; some appear to have the 
form of elongated needles. 

The first nephridium appears to be situated usually in segment viii. ; it was seen 
in vii. once, and once seemed to be in ix. The ante-septal portion of each is small ; the 
margin of the funnel projects on one side as a tag, from which long cilia wave down 
the lumen of the tube ; there are no outward cilia on the margin of the funnel. The 
post-septal portion is elongated, and in its anterior part is of a brownish colour (cf. L. 
subterraneus) ; the duct is shorter than the post-septal portion, and is directed obliquely 
downwards and backwards to the aperture ; just within the aperture the lumen is in 
sections seen to be dilated to form a small ampulla. 

The cerebral ganglion is indented posteriorly, while its anterior border is almost 
straight ; its lateral margins diverge somewhat posteriorly. The ganglion is about 
one-and-a-half times as long as broad. The ventral nerve-cord has " copulatory glands " 
associated with it in segments xiii., xiv., xv., and xvi. ; these closely embrace the cord 
laterally, but do not cover it on its dorsal surface. 

The testes are pear-shaped masses, in two groups, one on each side ; they are attached 
to septum \ y near the body-wall, laterally in the segment ; the lobes themselves may 
be either in segments x. or xi., according as the movements of the animal force them one 
way or the other ; the septum must therefore have considerable deficiencies. Sperma- 
tozoa usually occupy segments x. and xi., and may get forward into ix. 



44 



DR J. STEPHENSON ON 



The funnels are barrel-shaped, and comparatively short— -from one-and-a-half to two- 
and-a-half times as long as broad ; they have an everted lip. The vasa deferentia are 
contained in segment xii., and do not extend beyond this; they are long, very narrow, 
and closely coiled tubes ; their length was roughly estimated at about fourteen times that 
of the funnel. The male aperture has associated with it a large spherical penial gland ; 
the apertures lie in an area where the surface epithelium is low and cubical, and sharply 
marked off from the high clitellar epithelium around (PI. I. fig. 6) ; each aperture is at the 
anterior part of the glandular mass, and is lined by a continuation of the surface epithelium, 
with characters unchanged ; the short tubular passage thus constituted passes obliquely 
backwards for a short distance from the surface, and after receiving the termination of 
the vas deferens on its upper wall, ends behind by dividing into about three short branches 

oes 




?*■ 





tonvtfi. 



sj>.$. 



y ipk 



Yig. 5. 



a, anterior part of alimentary canal of Lumhricillus tuba, with appen- 6, another representation of the appearance 

dages, illustrating appearance of spermatheca in the living animal. of the spermatheca in the same species. 

Comm., communicating strand between septal glands and pharynx ; gl. , gland-cells round aperture of spermatheca ; 

oss., oesophagus;/;/;.., pharynx; spth., spermatheca; sp.gl., septal gland. 



(PI. I. fig. 6). The ]jenial gland is penetrated by the terminal part of the vas deferens, 
which enters the mass above, rather towards its lateral surface ; the gland has a 
muscular capsule, and is composed of elongated cells, which are individually very 
distinct — much more so, for example, than in my specimens of L. subterraneus ; these 
cells are arranged so that they radiate around the vas deferens in its course through 
the mass, and around the branching invagination of the external surface of the body 
(see description of male aperture above). 

The spermatheca, with their ducts, present a very characteristic appearance in the 
living animal, and may be designated as ' trumpet-shaped ' (figs. 5a, 5b). The ampulla 
is small, subspherical, thin -walled in its equatorial region, but with much thicker walls 
over its dome, i.e. around the situation of its communication with the oesophagus 
(PI. I. fig. 7); the difference being due to the different height of its epithelial 
lining. The duct is thick- walled, much longer than the ampulla, produced from the 



SOME LITTORAL OLIGOCHJETA OF THE CLYDE. 45 

ampulla without definite external demarcation between the two, and, as a rule, 
narrowing gradually towards the external aperture, so as to form an elongated and 
inverted cone. The lumen of the duct is narrow throughout ; the cells of which the 
inner layer of its wall is composed are covered by a conspicuous layer of muscular fibres, 
arranged longitudinally ; a cord of some hyaline matter ( ? coagulum) almost fills the 
lumen of the duct, in which may also lie a few spermatozoa. The duct is slightly 
invaginated into the cavity of the ampulla, though this is not evidenced externally ; 
there is consequently a circular trough around the ampullary opening of the duct, and 
in this trough the spermatozoa frequently lie coiled (PI. I. fig. 7). Surrounding the outer 
end of the duct is a fairly large lobulated gland, the cells of which are continuous with, 
and a modification of, the external epithelium round the aperture. Their inner ends 
are prolonged for a considerable distance within the muscular coat, as in other forms 
(PI. I. fig. 8) (cf. L. subterraneus, Encliytrseus albidus, L. viridis). The whole 
of the gland-cells are behind the level of septum f-, and the aperture is thus not in 
the intersegmental furrow, but posterior to this, on the anterior part of segment v. 

The characteristic trumpet or funnel- shape previously referred to is due to the 
gradual increase in the external diameter of the duct as it is followed inwards (the 
lumen is narrow and of the same diameter throughout) ; the margin of the funnel 
(fig. 5a) is the optical expression of the junction of the thick-walled duct with the thin- 
walled ampulla ; a small inner circle is the opening of the duct into the cavity of the 
ampulla. 

The clitellum extends over segments xii. and xiii. 

The intestine, in its anterior part at least, may be full of sporozoan parasites. 

I was for some time undetermined as to whether I should unite this form with 
Lambricillus (Pachydrilus) litoreus, Hesse (9). Though the descriptions agree in a 
number of points, they vary slightly in certain others, and considerably in the 
following: (i.) The length of the present form may be nearly twice that given for 
L. litoreus ; (ii.) the number of setse in the ventral bundles is four to six or seven in 
the present form, six to ten in L. litoreus; (iii.) the ccelomic corpuscles are more 
various in form, and contain a nucleus which is obvious in the fresh condition in the 
present species ; (iv.) the copulatory glands occur in segments xiii.-xvi., i.e. extend one 
segment farther back than in L. litoreus. 

The chief distinction, however, is in (v.) the spermathecse and their ducts; in L. 
litoreus the ampulla, according to the original description and its accompanying figure, 
is of comparatively large size, elongated in shape, with walls of the same thickness 
throughout, gradually merging into the duct, which latter is much shorter than the 
ampulla, and has two separate glandular masses at its aperture. In the present form 
the ampulla is small, subspherical, with extremely thin walls in its equatorial portion, 
thicker near its junction with the oesophagus ; there is, internally, a very sharp demar- 
cation between ampulla and duct, the latter being much longer than the ampulla, and 
being surrounded by a complete circle of large gland-cells at its aperture. The very 



46 DR J. STEPHENSON ON 

characteristic appearance of the spermathecae in the fresh specimen, Ivy which this form 
is easily identified, has been alluded to. 

Since the form and relations of the spermathecae and their ducts are among the 
most valuable characters for the discrimination of species in this group, it would seem 
advisable to separate this form under a special name ; the designation tuba is meant to 
refer to the trumpet-like appearance noted above. 

Lumbricillas viridis, n. sp. 

Found at Wemyss Bay, under stones below high-water mark, where fresh water was 
running to the shore. 

Length 1 inch ; the worm is stout, with a tapering anterior end. All but the 
anterior part of the body is of a green colour, due to the alimentary canal. They are 
active animals, and exhibit nematode-like contortions. Segments forty-five to forty-nine. 

The prostomium is bluntly conical. Numbers of hyaline cells are found in the 
superficial epithelium, arranged in fairly regular transverse rows over the whole extent 
of the body, most regular in its posterior part. 

The setse are in the usual four rows, two ventral and two lateral. They are straight, 
or almost straight, with a slight and usually gentle curve at their proximal end. Their 
points, especially on the anterior part of the body, are often blunt ; indeed the ends 
may be almost square, with rounded corners ; posteriorly sharper points are common. 
They are arranged fan- wise in the bundles ; all the setae of a bundle are not of the same 
length, but while in the ventral bundles the. inner (ventral) are shortest, in the lateral 
bundles the shorter setae are those on the dorsal side of the bundle. Thus in a bundle 
of seven ventral setae, in the anterior part of the body, the lengths of the individual 
setae, from the outer to the inner, were "105, '105, *105, "09, "085, '078, '066 mm. ; 
in another bundle, similarly, "112, -112, "105, "10, "095 ; in a lateral bundle situated more 
posteriorly, the lengths of the setse, beginning with the ventraimost seta of the bundle, 
were '090, *08, -075. 

The average length of the setae is therefore about one-tenth of a millimetre, longer 
ones occurring in the anterior part of the body ; their thickness is about "005 to "0045 
mm. The number of setae per bundle is five, six, or seven in the ventral series in front 
of the clitellum, three or four posterior to this ; there are no ventral setae on segment 
xii. The lateral setae are four or five, exceptionally three, in a bundle in front of the 
clitellum, two or three behind this. 

Notwithstanding the fact that the setae have been described above as straight or 
almost straight, I believe that the majority of them are to be considered as slightly, 
though very slightly, {-shaped. It is perhaps allowable to dwell for a moment on this 
point, since the genus Lumbricillus, in which the present form must, I think, be in- 
cluded, has |- shaped setae, and straight setae would be a curious anomaly. 

Fig. G, «, representing a bundle of seven setae from an anterior segment, and drawn 



SOME LITTORAL OLIGOCH^ETA OF THE CLYDE. 



47 



with the camera lucida, shows that the setae have in this particular instance blunt, in some 
cases almost square, ends, and are without trace of a distal curvature ; their proximal 
ends, however (with the exception of the innermost seta but one), show a gentle and 
gradual curvature, not the somewhat sharp curve seen in the genus Enchytrseus. The 
figure was drawn from a specimen in glycerin, and since the setal bundles are not per- 
fectly flattened, their two ends are at different levels and there is, perhaps, some optical 
distortion. This, however, will not apply to the single seta drawn in fig. 6, 6 : this was 
found in a section mounted in the usual way in balsam ; it belongs to segment iv. , and 
was drawn with the camera under an oil immersion lens. It will be seen that its distal 
portion is perfectly straight, and that it has a hooked inner end, the shape being that 




Fig. 6. — a, group of setfe from an anterior segment of Lumbrinlhis viridis. 

b, straight seta from segment iv. of a specimen of Lumbricillus viridis. 



characteristic of the genus Enchytrseus. Such "straight" setae therefore do occur in 
the anterior part of the body, in the present form. 

Fig. 7, «, shows two setae of a posterior bundle, drawn with a camera under an oil- 
immersion lens ; the ends of the setae are obliquely pointed as in Lumbricillus, not 
with the straight points seen in the genus Enchytrseus. 

I tried the effect of potash on the worms ; but they are so stout and firm that a 
strong solution of KOH has to act on them for hours before they collapse and flatten so 
as to present the setal bundles in one plane ; the setae then appear swollen and have 
evidently lost their true shape. Nevertheless, a double curvature of the Lumbricillus 
type, though not to be recognised in most of the anterior setae, is then, often fairly 
obvious in the setae of the posterior bundles. Thus fig. 7 , b, shows two setae of a posterior 
bundle in which this curvature is quite obvious ; and fig. 7, c, shows a fairly distinct 
double curve in a seta as far forward as segment viii. 

As supporting the view of the affinity of the setae with the Lumbricillus rather 
than with the Enchytrseus type, the fact of the difference in length of the setae of a 

TRANS. ROY. SOC. EDIN., VOL. XLYIII. PART I. (NO. 2). 8 



48 



DR J. STEPHENSON ON 



bundle may be mentioned. In the genus Enchytraeus the setse of a bundle are, as is 
well known, of equal length. 

Summing up, it may be stated, that while setse of the typical Enchytrseus type 
occur in the anterior part of the body, those of the posterior segments usually, and 
even some of those in the anterior segments occasionally, show a faint double curvature 
of the type found in the genus Lumbricillus. 

The ccelomic corpuscles are grey by transmitted light, flat, oval, or pear-shaped, and 
granular, with a distinct, clear nucleus. As seen in sections they are mostly *025 to '032 
mm. in their long diameter. 

The septa are thick and muscular, in accordance with the general build of the 





Fig. 1 .— a, distal ends, obliquely truncated, of two setse of a posterior bundle of a specimen of the same. 

b, two setae of a posterior segment of a specimen of the same, showing double curvature ; the 

more strongly curved ends are the proximal. The setae are swollen owing to the action of 
strong caustic potash solution. 

c, seta from segment viii. of a specimen of the same, showing slight double curvature ; swollen 

by the action of caustic potash. 

animal. It may be added that the retractor muscles of the pharynx are also very 
bulky, and that the muscular coat of the oesophagus is very well marked. 

The alimentary canal begins to be covered with chloragogen cells in segment iv. ; 
these cells become more numerous behind the septal glands, after which point the 
canal has a dark green colour. The oesophagus widens a little in segment vii., but 
there is no marked dilatation at any part of the tube. There are no peptonephridia. 
Septal glands occur in segments iv., v., and vi. ; the last two pairs are bulky, especially 
those of segment vi., which bulge backwards beyond the setse of segment vii. 

The dorsal vessel begins in segment xiii., and the blood is red. 

The nephridia are solid masses, with a very small anteseptal portion. The organs 
are elongated, ovoid, narrow and compressed laterally, the duct leading downwards 
from near the posterior end. 

The cerebral ganglion is comparatively small for the size of the worm, squarish in 



SOME LITTORAL OLIGOCH^TA OF THE CLYDE. 49 

shape, about as long as broad, and indented posteriorly. The ventral nerve-cord is 
characterised by the possession, through the whole of the anterior part of the body, of 
aggregations of cells in each segment which embrace the cord ventrally and laterally. 
These cells, though nowhere forming very prominent masses, correspond in appearance 
with those which constitute the ' : Bauchmarkdriisen " or copulatory glands of other 
forms. They are mostly, in sections, pear-shaped or spindle-shaped, with very feebly 
staining, slightly granular protoplasm and deeply staining nucleus ; they are of con- 
siderable size, and while they leave the dorsal side of the cord uncovered, frequently 
appear flattened over its sides and ventral surface, so as to give the appearance here of 
concentric layers. At the site of their occurrence the cord appears stalked in transverse 
section, being connected with the surface epithelium by prolongations of the cells. 

The testes are composed of six, seven, or more elongated club-shaped or pear-shaped 
masses on each side, springing from the ventral part of septem \^. Sperm morulse collect 
in segments x. and xi. The funnel is long, nine or ten times as long as broad ; the 
length, however, varies, and may appear to be only about seven times the breadth ; 
the funnels are bent on themselves ; the nuclei of the cells of which they are composed 
are, as usual, peripherally situated. The vas deferens is considerably coiled ; it does 
not, however, extend backwards behind the clitellum ; in diameter it measures '02 mm. 
There is a penial bulb of considerable size, its diameter being more than a quarter, 
nearly a third, of the diameter of the animal's body ; the bulb has a well-developed 
muscular covering ; the cells of which it is mainly composed are much elongated, 
radially arranged, the nuclei being crowded together peripherally ; the lumen is almost 
central. 

Ova are found in segments xii. and xiii. 

The spermatheccB are characterised by a spindle-shaped ampulla which com- 
municates with the oesophagus ; its external end is continued into the duct without 
evident demarcation. The duct is somewhat longer than the ampulla, and of about 
half the diameter of the latter ; it is surrounded by prominent gland- cells round its 
external aperture. The epithelium of the ampulla is columnar, but of very irregular 
height ; the nuclei of the cells are much elongated. In the duct, the epithelium is 
lower, and the nuclei are spherical ; the muscular coat of the duct is situated between 
and amongst the cells, the nuclei of the latter being all external to the muscular 
layer ; in the ampulla the muscular coat is, however, quite external to the epithelium. 
Both ampulla and duct are lined by a thick cuticular coat, continuous at the orifice 
with the very thin cuticle of the body-surface. The prominent collection of gland- 
cells round the duct near its external aperture consists of the epithelium of the duct, 
here much elongated and extending outwards far beyond the muscular layer of the 
duct. 

The clitellum is only peculiar in that it dies away gradually in front, without any 
definite line of demarcation. 

Sporozoa are present in the alimentary canal. 



50 DR J. STEPHENSON ON 

This species lias a very distinctive appearance, and can be immediately recognised 
by its stout form, active, wriggling, nematode-like movements, and especially by its green 
colour. 

The position of this species with regard to the setae has been discussed above, 
where it was shown that it holds, in this respect, an intermediate position. The lobed 
testes and the large compact penial bulb, however, determine the decision to place 
it in the genus Lumbricillus. The importance of this last feature, the penial bulb, 
in classification, has lately been insisted on by Eisen (4), who distinguishes two sub- 
families, Lumbricillinse and Enchytr winze, according to whether the penial glandular 
structures are or are not confined within a single bulb ; in the Lumbricillinse are 
included Lumbricillus, Marionina, Buchholzia, Stercutus, Bryodrilus, Henlea; in 
the Enchytraeinae, Enchytrseus, and Michaelsena. 

The ' Bauchmarkdrvisen ' are also a feature of the genus Lumbricillus. In the 
present species they are widely distributed, occurring throughout the whole of the 
anterior part of the body, but are small in size, and are only recognisable as such 
from the character of their cells as seen in sections. This comparatively undifferentiated 
condition may be contrasted with that which occurs in Lumbricillus subterraneus, 
where the glands, though few in number, are individually large and prominent. 

In a number of species in which the aperture of the spermathecal duct is surrounded 
by gland-cells, the muscular layer of the body-wall is continued between these cells, in 
the manner described and figured for Lumbricillus tuba (v. ant. ; also cf. Enchytrseus 
albidus, j>ost.) ; when, following the duct inwards, these gland-cells give place to the 
ordinary epithelium of the duct, we find, however, that the muscular layer is usually to 
be found outside the duct epithelium. In this species, however, the muscular layer is 
still to be found amongst and between the duct epithelium, having the same relations 
here as near the aperture. 

Enchytrseus nodosus, n. sp. 

Found at Wemyss Bay, near high-water mark, where fresh water ran to the shore. 

Length \ inch (8 mm.) ; small and thin, not tapering at either end. In colour the 
animals are intensely white over part, especially the posterior part, of their extent, but 
clear and transparent for the rest ; there may be only irregularly distributed white spots, 
or, as commonly, the posterior half of the body has intensely (opaque) white margins ; 
this opaque white coloration is due to aggregations of ccelomic corpuscles. Under the 
microscope the animal is, except for these aggregations, extremely transparent; and the 
clitellum, which extends over segments xii. and half of xiii. (to the level of the setae of 
the latter), is hardly less transparent than the rest of the body. Segments thirty-two 
to thirty-nine. 

The setse are of the straight type, with proximal hook (fig. 8, a) ; the sharpness of 
this hook varies, and in certain cases there is a faint indication of a double ({-shaped) 
curvature (fig. 8, b). Setae are absent (both ventral and lateral) in segment xii.; elsewhere 



SOME LITTORAL OLIGOCH^ETA OF THE CLYDE. 



51 



they are regularly two per bundle in both series throughout the body ; three were noted 
once only, one occasionally. 

There are no peptonephridia. Septal glands occur in the usual segments, those of 
iv. and v. being single, bulky, and situated doisally over the oesophagus ; in segment vi. 
there is a pair, elongated and extending a considerable distance backwards (PL I. fig. 9). 
The oesophagus, narrow in the region of the septal glands, is wider in segments vii.-ix. ; 
it narrows again in the genital region, and widens to become the intestine in segment 
xiv. The intestine is not constricted at the septa. The chloragogen cells are note- 
worthy ; they are large, with large refractile oil globules ; in sections they appear 
colour] ess, without brown or yellow granular pigment, and very markedly vacuolated, 
as if their contents had been dissolved out ; there are several or many large vacuoles in 
each cell. They are present here and there in segments v. and vi., though they can 
hardly be said to begin before vii. ; they are numerous and distinct in viii.-x., though 



i) 



Fig. 8. — a, seta of Enchytrxus nodosus. 

b, a seta of Enakytrmus nodosus showing 
a slight double curvature. 



not completely covering the oesophagus, and in particular they leave the tract of the 
dorsal vessel uncovered ; they are absent in the genital region, and begin again, thence- 
forward forming a complete investment of the intestine, in segment xiv. 

The dorsal vessel begins at the level of the setse of segment xiii., and bifurcates at 
the junction of the prostomium and first segment. The blood is colourless. 

The coelomic corpuscles are flat, circular, and, as measured in sections, are from "014 
or less up to '019 mm. in diameter. They are opaque by transmitted light, with a 
clearer nucleus ; the opacity is due to numerous refractile oil-like corpuscles crowded 
together so as to fill up the whole body of the cell. 

The nephridia (fig. 9) begin in segment viii. The ante-septal portion is of consider- 
able size, somewhat ovoid in shape, about one-third as long as the post-septal ; the open 
mouth of the tube is clothed with fine cilia, and its margin projects on one side as a short, 
overhanging process ; cilia beat in the tube in a downward direction, and the tube 
undergoes many windings before it reaches the level of the septum. A narrower neck 
connects the ante-septal with the post-septal portion ; the latter is elongated, narrow 



52 DR. J. STEPHENSON ON 

from side to side, with many and irregular windings of the lumen. The duct is stout, 
and leads downwards from the posterior end of the body of the organ ; in length it is 
about one-third of the post-septal. 

The cerebral ganglion is elongated, reaching as far as the level of the setae of 
segment ii. Its lateral margins diverge posteriorly, where it is indented at a blunt 
angle, as shown in PL I. fig. 9. The ventral nerve-cord shows small " copulatory glands " 
(Bauchmarkdrusen) in segments xiv. and xv. ; the cells of the glands embrace the cord 
laterally and ventrally, but not dorsally ; owing to the connection of these cells with 
the surface epithelium, the cord appears stalked in transverse sections at these situations ; 
there is externally a small transverse ridge opposite each gland. 

The testes are one on each side, in the usual position. The seminal funnels are 
four times as long as broad, of the usual cylindrical form, but a little narrower towards 
their attachment to the septum ; the lumen is obvious in the living condition, and in 
sections is seen to be not central but nearer the inner side ; the margin of the internal 
aperture, where the spermatozoa enter the tube, is everted, so as to form a small true 




Fig. 9. — Nephridium of the same. 

funnel, of a single layer of low columnar cells, perched on the cylindrical structure that 
usually goes by the name of ' funnel.' The vas deferens is thin, '0075 mm. in diameter ; 
it is but little coiled, reaches back to the level of segment xiii., and penetrates the 
penial gland nearer the outer than the inner side of the latter. The male aperture, in 
segment xii., consists of an invagination of the surface epithelium in a direction 
obliquely upwards and outwards (PI. II. fig. 10); this invagination, narrow from side to 
side, receives the end of the vas deferens, not at its upper extremity, but on its inner 
wall ; the invagination and its immediate neighbourhood are covered by a distinct and 
fairly thick cuticle. The penial gland (PI. II. fig. 1 0) is a single ovoid mass on each 
side, consisting of much-elongated cells with nuclei near their internal ends ; their 
external ends form the inner wall of the invagination previously referred to ; the gland 
has a distinct though feeble muscular covering, and muscular strands pass obliquely 
upwards from its surface towards the lateral body-wall ; other strands, passing over it 
obliquely from the ventral to the lateral body- wall, appear, so to speak, to bind it down. 
The spermathecse consist of ampulla and duct, both of which present, especially in 
sections, a peculiar appearance, from the fact that the cells of which they are composed 
are irregular in size, shape, and disposition, and thus are very far from forming a regular 



SOME LITTORAL OLIGOCH^ETA OF THE CLYDE. 53 

epithelial lining (PI. II. fig. 11). The ampulla communicates as usual with the oeso- 
phagus; in the living specimen it appears somewhat irregularly spherical, with projecting 
bosses. In sections the cavity is irregular in shape, possessing a few small saccular 
diverticula ; the cells composing the wall are comparatively few, large in size, irregular in 
shape and arrangement, and do not form a regular epithelial layer. The spermatozoa 
penetrate between the cells, and the appearances in sections seem to show that they burrow 
into the cells themselves ; they also seem to penetrate the wall, and are found lying on 
the outer surface (PL IT. fig. 11). The ampulla possesses no muscular covering. The 
duct leads obliquely forwards to the exterior ; it is a little longer than the ampulla, and, 
like the latter, appears in the living condition to be studded with small rounded projections, 
smaller, however, than those of the ampulla ; round the external aperture these projecting 
cells are again larger, and form a distinct rosette of glands. The statements as to 
irregularity of shape and disposition of the cells, and absence of a definite epithelium, 
already made for the ampulla, hold also for the duct ; so that, except near the external 
aperture, it is difficult to follow the lumen in sections. The cuticle of the surface is 
invaginated for some little distance at the external aperture, and muscular fibres con- 
tinuous with those of the body- wall are conspicuous among the cells of the duct ; but 
these latter are not continued on to the ampulla. 

The clitellar cells are entirely wanting over the mid-ventral region of the body, 
along an area whose breadth is the distance between the male apertures of the two 
sides (PL II. fig. 10). 

The species to which the present form shows most resemblance, at least externally, 
seems undoubtedly to be Enchytrseus argenteus, Michaelsen (11). The same opaque 
white colour characterises both, and in both is due to the same cause — the presence of 
opaque granular coelomic corpuscles ; further, owing to the varying aggregations of these 
corpuscles in different parts of the body, the whiteness is not uniform in either. 

But, apart from such indifferent features as the colourless blood and the absence of 
peptonephridia in both (probably, since Michaelsen does not mention these structures), 
the resemblances seem to end here. The differences in size (E. argenteus, 2*5-5 mm., 
present form 8 mm.), in number of segments (E. a. twenty-three to thirty, present form 
thirty-two to thirty-nine), and in number of setse per bundle (E. a. two or three, present 
form regularly two), are not very great ; the chief differences are those of the cerebral 
ganglion, nephridia, seminal funnel, and spermathecse. 

The cerebral ganglion has a rounded posterior end in E. argenteus, while in the 
present form it is indented posteriorly ; the seminal funnel is short, somewhat longer 
than broad in the former, while it is four times as long as broad in the latter. The 
ampulla of the spermatheca is of an inverted pear-shape in the former, and the duct 
is simple (without gland-cells) ; the peculiarities of these structures in the present form 
have been described above. The nephridia of E. argenteus are not constricted at their 
passage through the septa, and the lumen forms a small number of regularly arranged 
and consecutive loops in the post-septal portion ; in the form under description there is 



54 DR J. STEPHENSON ON 

a constriction at the septum, and the lumen undergoes many and irregular windings 
in the post-septal portion. I propose, therefore, to consider the present form as a new 
species, under the name EnchyPraBUS nodosus* 

It is interesting to note here again, as has already been done from the other side in 
the case of Lumbricillus viridis, indications of transition between the genus Enchy- 
tra&us and the Lumbricilline group. In the present case these are (l) the seta shown 
in fig. 8, b, with its double curvature, as opposed to the straight setae of Enchytraeus ; 
(2) the presence of small ' copulatory glands ' in segments xiv. and xv. ; and (3) the 
definite and single penial bulb. I have already referred to the importance assigned by 
Eisen to this structure, and to the fact that this author distinguishes two sub-families, 
the Lumbricillinse and Encliytrasinse, according to the presence of a single penial bulb, 
or its substitution by a number of separate aggregates of gland-cells. 

It seems doubtful whether the presence or absence of a penial bulb is of sufficient 
importance to serve as a basis for the distinction of sub-families, or even, perhaps, of 
genera. And it is interesting in this connection to compare Michaelsen's figure (10) 
of the structures round the male genital aperture in Enchy trams mobii ( = olbidus), which 
shows that there is there a true " penial bulb " surrounding the end of the vas deferens, 
such as is met with in Lumbricillus ; it is, however, of comparatively small size, and 
there are in addition separate aggregates of gland-cells on each side of the bulb. In 
other words, there is a condition intermediate between, or representing a combination of, 
those described by Eisen as characteristic of his two sub-families. 



Enchytraeus dubius, n. sp. 

Found under stones, between tide-marks, at Wemyss Bay. While the majority of 
specimens of other species of Enchytraeids were sexually mature from May to July, in 
this case the greater number of specimens were without sexual organs. 

The animals showed a great tendency to curl up. In length they were half an inch 
(12 mm.) or less. In colour they were whitish ; examined with a lens they were only 
moderately translucent under pressure, and showed a considerable amount of white 
opacity in the middle region of the body along the borders of the alimentary canal, due 
to aggregations of coelomic corpuscles and chloragogen cells. The clitellar region was 
no more opaque than the rest of the body. 

Segments, forty-four. Prostomium rounded or very bluntly conical, with minute 
secondary projections. Head-pore between prostomium and first segment. 

The setae are in four rows, two ventral and two lateral. With very rare exceptions, 
there are two setae per bundle throughout, except that ventral setae are always absent 
in segment xii. Botli ventral and lateral setae are of the same shape, straight, with a 

* The E. porvuhis of Friend (6, 7), is doubtfully identified by Michaelsen (11), with E. argenteus. The data do 
not permit a detailed comparison of E. parvulus with the present form ; but the two would seem to differ, at any rate, 
in the numbers of the setse and shape of the cerebral ganglion, and less markedly in size and number of segments. 



SOME LITTORAL OLIGOCHOLIA OF THE CLYDE. 55 

proximal curve, thickest in the middle (v. fig. 10, a). In length, the setae vary consider- 
ably, from "045 to '072 mm. ; the average is about '06 mm. There is no constant 
difference between the lengths of ventral and lateral setae ; nor between those of the 
anterior and posterior regions of the body, except that the average length of the posterior 
is perhaps a little less than that of the anterior setae. 

The setae also vary considerably in thickness, viz. from '006 mm. to '0045 mm. 
The sharpness of their points varies, probably to some extent at least with age ; newly 
formed setae (in which the basal curved portion is not yet present) have sharp points, 
while in others they may be quite blunt, almost truncated. 

Many setae show a refractile, elongated, sometimes spindle-shaped body in their 
centre, about the middle of their length (in glycerin preparations) ; the appearance is 
possibly due to some separation of the component fibrils (fig. 10, a). The setae appear to 
be shed periodically ; there may occasionally be seen two newly forming setae with an old 
one, in the same bundle, and thus there is presented the appearance of three setae per 
bundle. 

The alimentary tract shows no demarcation into separate regions between the 



^ 



<i=- 



Fig. 10. — Setfe of Enchytrseus dubius : a, pointed ; b, blunt, 
with refractile appearance in its centre. 

pharynx in the second and third segments and the intestine, which begins suddenly in 
the fourteenth. The septal glands in segments iv. and v. are of moderate size, those in 
segment vi. are large ; the individual cells composing the glands are visible in the fresh 
state. There are no peptonephridia. Chloragogen cells begin in segment vii., or there 
may be a few in segment vi. ; they are comparatively few and discrete up to xi., absent 
in xii. and xiii., numerous and close-set from xiv. onwards till near the posterior end, 
where they are fewer, and finally absent. The cells are of large size, with prominent 
oil-drops ; in sections they are tall, elongated vertically to the alimentary wall, and 
present numbers of vacuoles. The alimentary canal is attached by stout strands to the 
ventral body -wall in each segment. 

The dorsal vessel may begin in segment xiv., or at the posterior boundary of segment 
xii. ; it bifurcates in the prostomium. The ventral vessel is formed about the level of 
septum § by the union of the two terminal branches of the dorsal vessel. There are 
four commissural loops in the anterior part of the body ; these are contained mostly in 
the third and fourth segments, but their exact position is not, apparently, always the 
same. The blood is red. 

The ante- septal portion of the nephridia is small ; very fine cilia are attached to 
the rim of the funnel, and longer cilia beat down the lumen ; the post-septal is of 

TRANS. ROY. SOC. EDIN, VOL. XLVIII. PART I. (NO. 2). 9 



56 



1)R J. STEPHENSON ON 



a stout ovoid shape ; tlie duct or terminal portion of the nephridium comes from the 
posterior end of the post-septal, is remarkably stout, constricted at the orifice, has a 
vertical course, and is equal to the post-septal in length. 

The ccelomic corpuscles are numerous, and have the form of large flat discs, irregularly 
circular or oval in shape, very coarsely granular, with a small nucleus which is not 
obvious in the fresh state but is visible in stained preparations. By transmitted light 
the corpuscles are grey. In diameter, they were estimated at about '045 mm. in the 
living animal ; but in sections they are about '03 mm., the largest being 033. As in 
so many species, they may be discharged in large numbers from near the anus under 
pressure ; after being shed they become regularly circular in outline. 




Fig. 11.— Outline of male organs in segment xi. of a specimen of the same (sketch from the living animal) : 
/., funnel ; t., testis of one side ; t. 1 , testis of the other side, faintly seen. 



The cerebral ganglion is nearly twice as long as it is broad, is deeply indented 
posteriorly, and extends back as far as the level of the setse of the second segment. 

The testes, on the posterior face of septum j-f , are large, and resemble somewhat 
those found in the genus Lumbricillus. They have a limited origin from the septum ; 
from this limited origin there springs an elongated, coiled, or bent cellular cord, which 
may swell to an irregular bulky mass, and gives off, near or at some distance from its 
base, two or three branches, of the same character, and, it may be, almost of the same 
size as itself (fig. 11, PL II. fig. 12). Sperm morulse may be present in all segments 
from vi. to xiii. 

The funnels are comparatively small, about four times as long as broad, narrower 
towards their attachment to the septum. The vas deferens is long, thin, coiled, in 
segment xii. The pe?iial gland is not large ; its peculiarity is that it is bifid internally ; 
thus in a series of longitudinal sections it is first met with as a single mass (fig. 12, a), 



SOME LITTORAL OLIGOCH^TA OF THE CLYDE. 



57 



while, nearer the middle line, it is completely double (fig. 1 2, b). It is attached by two 
thick strands, composed of cells with large oval nuclei, to the dorso-lateral body-wall. 
A portion of the strands, which are so disposed as to be one anterior and one posterior, 
passes internal to the gland ventrally to be inserted into the ventral body-wall ; the 
glands are thus to some extent bound down by the strands. Dorsally the strands 
split up and radiate to their attachments. 

The ovaries have the usual position. Ova are found in segments xii. and xiii. 

The spermathecte in segment v. are not large, and have the form of an elongated 
spindle, somewhat bent on itself. The communication with the oesophagus is narrow. 
Gland-cells are disposed in radial masses round the external aperture. There is no 
distinction of ampulla and duct to be made out in the entire animal ; sections, however, 



vi. 





a Fig. 12. b 

a, longitudinal section through outer part of penial b, a similar section, a little internal to the previous 

gland of the same. one. The penial gland here appears double. 

m., muscular (cellular) strands attaching penial gland to body-wall ; m. 1 , muscular covering of gland itself ; v.d., vas 

deferens before entering the gland. 



show a small ampulla with thin walls, and a much longer and thicker walled duct, with 
a fairly wide lumen (PL II. fig. 13). The walls of the duct are composed of columnar 
cells, and appear markedly transversely striated in the living condition (PL II. fig. 14). 

It is possible that the ampulla would be relatively larger if it were swollen by 
spermatozoa ; none of the organs in my preparations, however, contain any. 

Copulatory glands (" Bauchmarkdriisen ") are well marked in segment xv. ; the 
mass of cells closely invests the cord, and projects upwards on each side above the level 
of the cord ; there is, in longitudinal sections, a small papilliform projection of the 
surface of the body at the level of the middle of the gland. 

In xiv. the copulatory gland is smaller, but still projects on each side above the 
level of the cord. In segments anterior to this, from xiii. to ix. , the cells around the 
cord in the posterior part of each segment appear to be of the same nature, as evidenced 
either by the papillary projection on the surface or by the fact that the cells penetrate 
the muscular layers of the body -wall to become continuous with the surface epithelium. 



58 DR J. STEPHENSON ON 

I have referred in the Introduction to the several Lumbricilline features exhibited 
by this worm. 

Enchytrseus sabulosus, Southern. 

This species was discovered by Southern (12) in Dublin Bay, living under stones 
and amongst the gravel at high- water mark. This is apparently the only record of its 
occurrence ; I therefore give a few notes on Scotch specimens which I believe to be 
identical with it. 

The worm was found at Wemyss Bay, under stones near high-water mark, at a spot 
where fresh water w T as running to the shore. 

Length about f inch (18 mm.); colour white. Anterior end tapers somewhat, 
posterior end blunter. Prostomium bluntly conical. Segments forty-six to forty-nine. 
Clitellum on segments xii.-xiii. 

Setse of the form usual in the genus, in length '071 to "088 mm. Their peculiarity 
lies in the number per bundle — with few exceptions three in the ante-clitellial, two in 
the post-clitellial bundles ; occasionally there are three setae in a post-clitellial bundle, 
and young replacing bundles are sometimes seen near the functioning bundles. In 
segment xii. there are no ventral setse, and the dorsal setse are in this segment two per 
bundle. The fact that in my specimens the post-clitellial bundles have only two setae 
is the most important difference from Southern's description ; for he states that the 
number is regularly three throughout the body. 

Septal glands bulky, the last pair being the largest. Peptonephridia (PL II. 
fig. 15) as small hollow tubes, bent once or twice, or slightly coiled, extending backwards 
as far as the first pair of septal glands, and opening anteriorly close together into the 
pharynx on its dorsal wall. I have no note of any special peculiarity of the cliloragogen 
cells, which begin in segment vii. ; Southern considers their large size and their 
oil-drops to be of value as a specific distinction. 

The dorsal vessel begins in segment xv. (junction xvi. and xvii., Southern). The blood 
is colourless. The coslomic corpuscles are irregular, ovoid or pear-shaped, granular, with 
a clear nucleus. The nephridia are as described by Southern ; they begin in segment vii. 

The cerebral ganglion is one-and-a-half times as long as broad (PL II. fig. 15), its 
sides nearly parallel, its posterior end rounded, not indented. Two small dark spots 
may be seen on it, as in specimens of E. albidus ; but they are not so conspicuous in 
the present form. The ventral nerve-cord shows ganglionic swellings in segments ii., iii., 
and iv. ; thereafter the swellings are slight or absent, and the cord, as seen in the living 
animal, is of the same thickness throughout. 

The sperm morulse may bulge forwards as far as the level of the setse of segment 
viii. The sperm funnels, about four times as long as broad, are as described by 
Southern. The vas deferens is a stout tube, not much coiled, extending back as far 
as segment xviii. The glands round the male aperture are constituted by a number of 
separate aggregations of cells, and do not form a single penial bulb. 



SOME LITTORAL OLIGOCPLETA OF THE CLYDE. 



59 



the general form is 



The spermathecal apparatus is represented in PI. II. fig. 15 
somewhat similar to that illustrated in Southern's figures ; the ampulla, comparatively 
small, spherical or slightly elongated, is smaller, and the duct, tuberculated as in E. 
albidus, is thicker than there depicted. 



Enchytrseus albidus (Henle). 

This worm has frequently been described, but under a very large number of different 
names. According to the synonymy given by Michaelsen (11), thirteen authors have, 
in eighteen papers, given to this animal five generic and twelve specific names. Perhaps 
the fullest account of the worm is that given by Michaelsen (10) in 1886, in his thesis 
Untersuchungen iiber Enchytrseus Mobii, Mich., und andere Enchytrwiden ; Goodrich 



(1 



I 




Fig. 13. — a, two setae of Enchytrseus albidus. 

b, a bundle of sete of Enchytrceus albidus, 
showing two immature replacing setae, 
along with two fully formed setae which 
are destined to drop out. 



(8) has more recently, under the name E. hortensis, described a form which Michaelsen 
considers identical with the above, and has paid special attention to the nephridia and 
coelomic corpuscles. A few remarks, chiefly in regard to points in which the Millport 
specimens vary from the above descriptions, will therefore be sufficient. 

The worms were found about high-water mark in places where fresh water was 
running to the shore ; it was common under stones, and also among the roots of plants. 
In length specimens are f to 1^ inches ; they are comparatively stout, the anterior end 
tapering, the posterior blunter ; they are whitish in colour, fairly transparent, and easy 
to examine by the microscope in the living state. Segments fifty-two to sixty-six. 
The clitellum occupies segments xii. and xiii. The animals move by crawling, or at 
times by wriggling ; they often throw themselves into nematode-like contortions ; at 
rest, in a dish, they curl themselves up. 

As to the setae, previous descriptions, and fig. 13, «, will be a sufficient guide as 
to their shape. There are no ventral setse in segment xii. ; they are more numerous in 
the anterior than the posterior segments, the numbers being three to five (commonly 
four) in the ante-clitellial ventral bundles, two to four (commonly three) in the 
post-clitellial ventral bundles ; for the lateral bundles the numbers are two to four 



60 DR J. STEPHENSON ON 

ante-clitcllial, and two or three post-clitellial. In length the sette of the posterior 
part of the body appear to be, on the average, rather longer than those of the 
anterior ; the longest may measure up to "1 mm. The bundles are replaced during 
the life of the animal (fig. 13); thus there may appear to be eight setse in a bundle, 
but of these one group of four will be immature, wanting the proximal hooked 
end. 

The alimentary tract (PI. II. fig. 16), with its appendages (septal glands, pepto- 
nephridia), corresponds with previous descriptions, except that I have not found, as 
Goodrich states for his form, the last pair of septal glands smaller than the others. 
The dorsal vessel may take its origin from the intestinal sinus in segments xiv., xv., 
xvi., xvii., or xviii ; the lateral commissures may be four in number, in segments ii., iii., 
iv., and v., but I am not satisfied that this number and arrangement are constant. 
The blood is colourless. 

One kind of lymph corpuscles only is mentioned by Michaelsen — flat, oval, or 
pear-shaped cells, nucleated, with a large nucleolus ; Goodrich mentions three kinds, 
and gives a detailed description of each. In my specimens I noted two forms of 
ccelomic corpuscles — one granular, flat, irregularly pear-shaped or oval, and nucleated, 
corresponding to Michaelsen's description ; the other spherical or irregular, not 
flattened, homogeneous, more refractile than the first type, apparently not nucleated 
and not so numerous as the first kind. These may perhaps correspond to the first type 
of ccelomic corpuscle described by Goodrich. 

The cerebral ganglion lies in segment i., attached to the dorsal wall of the buccal 
cavity ; it is one-and-a-half times as long as broad when the head is extended ; its 
posterior border is convex or flattened. The two authors already quoted both found 
it to be slightly indented behind. A pair of spots occur near the posterior margin of the 
ganglion, dark by transmitted light. They were often very large and conspicuous, 
and sometimes contained a few refractile particles besides the usual granular matter 
of which they seemed to be made up. They were not always quite symmetrically 
placed (PI. II. fig. 16). 

There is a well-marked tubular cavity dorsally in the substance of the ventral 
nerve-cord all through the clitellar region, and for some distance in front of this ; 
it splits up into several smaller tubes in the region of the last septal glands, and 
some of these tubes can be followed for some distance farther towards the head. 

'Y\\e funnels of the vasa deferentia vary much in shape; when the animal stretches 
itself out, they may be seven or eight times as long as broad ; ordinarily they are perhaps 
about five times, and sometimes may appear as little as three times as long as broad. 
The vasa deferentia, may extend backwards as far as segment xxi. The vesiculse seminales 
are constituted by a bulging forwards of septum \\ ; thus masses of spermatozoa are 
seen to surround the oesophagus in segments x. and xi., or ix., x., and xi. 

The shape of the spermatheciv deserves mention, since it differs from that described 
by the two authors previously quoted. Thus Michaulsen's figure shows the cavity of 



SOME LITTORAL OLIGOCH^ETA OF THE CLYDE. 61 

the spermatheca as squarish, with no special bulging anywhere ; while, according to 
Goodrich, the oesophageal and external openings of the spermatheca are about at the 
same level, but the cavity of the ampulla is produced backwards into a large posterior 
sac. In my specimens the spermathecse appear in an early stage of their development 
as simple tubes, not dilated anywhere, passing obliquely backwards from their external 
opening between segments iv. and v.. to the oesophagus. In the fully formed organ the 
ampulla is large, ovoid in shape, with long diameter antero-posterior ; it fills up the 
space on each side between the oesophagus and body-wall. It opens into the oesophagus 
near its posterior end, the aperture of communication being ventrally placed with regard 
to the cavity of the ampulla. In front the ampulla passes into the duct, the boundary 
between the two being, in the fully dilated condition of the ampulla, quite sudden. 
The duct is about as long as the ampulla, and forms a stout tube, straight, or more 
usually, in the contracted condition of the animal, somewhat bent ; the outline of the 
tube is irregular, appearing to be studded with small excrescences ; these irregularities 
are due to the projection of the cells of which it is composed beyond the muscular layer. 

Some ciliated parasites were seen on one occasion in the body-cavity. 

I think there is no doubt that this worm is most suitably included under E. albidus, 
in spite of a few divergences from previous descriptions. These divergences seem to be 
the following : — (i.) Extent of clitellum ; this Michaelsen gives as half xi. to half xiii., 
while his figure shows it as extending nearly to the anterior border of xi., and leaving 
a large part of xiii. unincluded ; (ii.) the lymph-corpuscles (v. sup.) ; (iii.) the dark spots 
on the cerebral ganglion (which may, however, merely have gone unrecorded) ; (iv.) the 
difference in the canals of the ventral nerve-cord in Michaelsen's description and mine ; 
(v.) the difference in the shape of the spermathecaa. To these may be added the fact 
that I have not noted in my specimens collections of sensory cells near the apertures of 
the spermathecse, as figured by Michaelsen. 

Fridericia bulbosa (Kosa). 

This species is widely distributed, and has recently been recorded from Ireland ( 1 4). 
It appears, however, to be somewhat variable, and different authors have given different 
descriptions of, for example, the shape of the cerebral ganglion, the form of the pepto- 
nephridia, and the ducts of the spermathecae with regard to the presence or absence of 
gland- cells round the orifice. A brief account of the features in which, from the 
descriptions of previous observers, some amount of variation appears to have been 
established, may therefore be of interest. 

The first point is the habitat of the Millport specimens. They were found under 
stones, between tide-marks, at Balloch. The species lives, according to Michaelsen (10), 
in rotten wood or damp leaves ; indeed the genus Fridericia as a whole " is terrestrial, 
and found in the driest localities" (Beddard, 1, p. 312), a fact which Beddard brings 
into relation with the occurrence of dorsal pores in the genus. 



62 



DR J. STEPHENSON ON 



As to its anatomical features, the worm was 3 to |- inch (8 to 12 mm.), in length, fili- 
form, white in colour, very sluggish. Prostomium short, rounded ; head-pore visible as 
a somewhat elongated slit, at the junction of prostomium and first segment; clitellum 
extending over most of xii. and half or more of xiii. ; segments thirty-eight to forty-five. 
Dorsal pores from vii. onwards, some little distance behind the septa ; with two cells in 
relation to each, granular and with large nuclei, one anterior and one posterior. 

The setx are absent in xii. ; form, numbers, and distribution as previously recorded. 
The larger (outer pair) setae are comparatively short in the first segments ("047 mm.), 
and their length increases towards the clitellar region, near which it attains a first 
maximum (066 mm.) ; diminishing in the middle region of the body ("048 mm.), the 
length again increases, and reaches a second maximum, higher than the first ("075 mm.), 
near the posterior end. Their thickness varies a little ; it is often about a tenth of 
their length, or even more, i.e. "0044 to '0057 mm. (cf. fig. 14, a). 




Fig. 14. — a, large setae of Fridericia bullosa. 

b, a group of setfe of the same, showing 
relation of smaller setse to the larger. 



The smaller setae, included between the larger in the anterior part of the body, 
have their outer ends on a level, or almost so, with the ends of the large setse, but 
their hooked inner ends are at a more superficial level. In length they are about two- 
thirds the size of the larger, i.e. '03 to '04 mm. (cf. fig. 14, b). 

The peptonephridia enter the oesophagus in segment iv., and extend back to the 
level of the setse of v. They are not, in my specimens, branched or expanded at 
their ends. 

The dorsal vessel originates in segments xix., xx., or xxi. The blood is colourless. 

The ccelomic corpuscles are all of one kind, large, flat, oval discs, granular, nucleated, 
up to '022 mm. in diameter. The nephridia have the characters given in previous 
descriptions, the ante-septal portion being comparatively large, about one-third to one- 
fourth the length of the post-septal. The cerebral ganglion is twice as long as broad, 
with a rounded posterior end ; it extends back a considerable distance into segment ii. ; 
its lateral margins converge somewhat towards the front. The ventral nerve-cord has 
copulatory glands (' : Bauchmarkdriisen ") associated with it in segments xiii. and xiv. 



SOME LITTORAL OLIGOCHSETA OF THE CLYDE. 63 

This species shows the commencement of definite sperm and ovisac ; septum -^ is 
markedly bulged backwards, so as to reach what would normally be the hinder limit 
of the next posterior segment, and the developing sperm-morulge, contained in the 
sac so formed, do not pass beyond it ; similarly the ova are contained in a posterior 
bulging of septum ^§. 

The funnels are between two and three times as long as broad; the vas deferens 
is a coiled narrow tube, confined to segment xii. ; there is a well-marked penial bulb 
immediately on the inner side of the termination of the vas deferens at the male aper- 
ture ; the vas can thus hardly be said to perforate the bulb (PI. II. fig. 17). The 
spermatheca and its duct have the form and relations described by previous observers ; 
there are, in my specimens, no glands round the duct or its aperture. 



LITERATURE. 



(1) Beddard, F. E., A Monograph of the Order of Oligochseta, Oxford, 1895. 

(2) Benham, W. B., "Notes on some Aquatic Oligochaeta," Quart. J own. Microsc. Science, N.S., vol. 

xxxiii., 1891. 

(3) Claparede, E., " Etudes anatomiques sur les Annelides, Turbellaries . . . observes dans les Hebrides," 

Mem. de la Soc. de Physique et d'Hist. Nat. de Geneve, tome xvi., part i., Geneva, 1861. 

(4) Eisen, G., "Enchytrseids," Harriman Alaska Expedition, vol. xii., New York, 1904. 

(5) Evans, W., "The Oligochseta of the Forth Area," Proc. Roy. Phys. Soc. Edin., vol. xviii, No. 2, 

1910. 

(6) Friend, H, "A new British Worm," Zoologist, ser. 4, vol. i., 1S97. 

(7) Friend, H, " Studies in Irish Enchytrseids," Irish Naturalist, vol. xi., 1902. 

(8) Goodrich, E. S., "Notes on Oligochsetes, with the description of a new species," Quart. Journ. 

Microsc. Sci., N.S., vol. xxxix., 1896. 

(9) Hesse, R., " Beitrage zur Kenntnis des Baues der Encliytraeiden," Zeit. f. wiss. Zool., vol. lvii., 

part i., 1893. 

(10) Michaelsen, W., Untersuchungen iiber Enchytrceus Mobii, Mich., und andere Ewhytrxiden, Kiel. 

1886. 

(11) Michaelsen, W., "Oligochpeta," in Das Tierreich, Berlin, 1900. 

(12) Southern, R., "Notes on the genus Enchytrxus . . .," Irish Naturalist, vol. xv., 1906. 

(13) Southern, R., "Oligochseta of Lambay," Irish Naturalist, vol. xvi., 1907. 

(14) Southern, R., " Contributions towards a Monograph of the British and Irish Oligochseta," Proc. 

Roy. Irish Acad., vol. xxvii., Sect. B, No. 8, 1909. 

(15) Vicjdovsky, F., "Note sur le Pachydrilus subterraneus, n. sp.," Rev. Biol. Nord France, vol. i., 1889. 

(16) Vejdovsky, F., System und Morphologie der Oligochxten, Prag, 1884. 



EXPLANATION OF PLATES. 



Fig. 1. Diagram of the vascular system of Tubifex costatus. 
Fig. 2. Part of the anterior end of Marionina semifusca. 

Amp., ampulla of spermatheca; comm., communicating cord from septal glands to pharynx; d., 
duct of spermatheca; gl., gland-cells round spermathecal aperture; n., nephridium of segment 
v.; as., oesophagus; s^s®, septal glands of the fourth to the sixth segment; ph., pharynx; 
sp., septum |^. 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 2). 10 



64 DR J. STEPHENSON ON 

Fig. 3. Part of the anterior end of Lumbricilhis subterraneus. 

Comm. 1 , communicating strand from septal gland of the fifth to that of the fourth segment; spth., 
spermatheca (duct and ampulla not distinguishable). Other references as for fig. 5. 
Fig. 4-. Transverse section of ventral nerve-cord and ' copulatory gland ' of Lumbricilhis subterraneus. 
x640. 
b. bridge of deeply staining tissue connecting the two sides of the gland dorsal to the cord \ c, 
central canal of the nerve-cord; cm., circular muscular layer; ep., surface epithelium, in 
which cell outlines are indistinguishable ; gl., a cell of the copulatory gland, the body of the 
cell staining equably; gl. 1 , another gland cell, the cell body showing a reticular structure and 
staining very slightly; l.vi., longitudinal muscular layer ; n., fibrous part of nerve-cord, below 
and to the right of which are seen the granular nuclei of the ganglion cells ; st., 'stalk 'of 
attachment of copulatory gland to the surface epithelium. 
Fig. 5. Longitudinal section through spermathecal apparatus of Lumbricillus subterraneus. x 220. 

.1/., alimentary canal; amp., ampulla of spermatheca; ap., aperture of its duct; c, circular 
muscular layer of body-wall; d., duct of spermatheca; ep., surface epithelium; gl., gland- 
cells near aperture of spermatheca ; I., longitudinal muscular layer of body-wall ; m., muscular 
layer of duct of spermatheca ; m. 1 , the same layer between the cells, where these are elongated 
and glandular; s.gl., septal gland; spz., spermatozoa in ampulla of spermatheca. 
Fig. G. Longitudinal section through male aperture of Lumbricillus tuba. x 250. 

Gl., clitellar epithelium; ep., lower, more deeply staining epithelium round male aperture; m., 
muscular capsule of penial gland; m.\ muscular strand of attachment, cut through; v.d., vas 
deferens ; <$ , male aperture. 
Fig. 7. Longitudinal section through ampulla and first part of duct of spermatheca of the same, x 250. 

Amp., ampulla; d., duct; oes., oesophagus; s.gl., septal gland of fourth segment. 
Fig. 8. Longitudinal section through aperture of duct of spermatheca of the same, to show the continuity 
of the surrounding gland-cells with the cells of the surface epithelium and of the duct, and their relation to the 
muscular layer. x 250. 

M., muscular layer clothing duct, continued ventrally between gland-cells; s.gl., septal gland. 
Fig. 9. Anterior part of the body of Enchytreeus nodosus, semi-diagrammatic. 

Amp., ampulla of spermatheca ; e.g., cerebral ganglion ; conn., connection between septal glands 
and pharynx ; conn}, that between the glands of segments v. and iv. ; d., duct of spermatheca ; 
.'/., gland-cells round aperture of spermatheca ; m., muscular strands attaching pharynx to 
body- wall ; oes., oesophagus; ph., pharynx; s.gl. 1 ' 3 , septal glands of segments iv.-vi. ; vac, 
vacuole-like appearances in posterior part of cerebral ganglion. 
Fig. 10. Transverse section through male aperture of the same. x 550. 

CI., clitellar epithelium; cut., cuticle; ep., lower, non-glandular epithelium over area around and 
between male apertures; invag., the invagination representing the male aperture, at the 
place where it receives the termination of the vas deferens ; m., muscular strand from penial 
bulb to body-wall; m. 1 , muscular covering of penial bulb; p., cells of penial bulb; v.d., vas 
deferens before penetrating penial bulb; v.n.c, ventral nerve-cord. 
FlG. 11. Longitudinal (somewhat oblique) section through spermatheca of the same. x 640. 

Amp., cells composing wall of ampulla; cm., circular muscle layer of body-wall; cut., cuticle; ep., 

surface epithelium; g., gland-cells near aperture of spermatheca ; l.m., longitudinal muscle 

layer of body-wall ; per., peritoneal cells ; s. gl., septal gland of segment iv. ; spz. 1 , spermatozoa in 

ampulla ; spz?, spermatozoa between cells of wall of ampulla ; spz, a , spermatozoa outside ampulla. 

Fig. 12. Longitudinal section through segment xi. of a specimen of Emhytrcnis dubius. x 250. 

B.-v,, a blood-vessel; Corp., coelomie corpuscle; dis., dissepiment li ; /., sperm funnel; sp., mass 
of developing sperm cells ; t., testis at its attachment to septum }^ ; t. 1 , other portions of the 
brandling testis. 
I'm;. 13. Longitudinal section through spermatheca of Enchyirxus dubius. x 250. 

Ami)., small dilatation, with thinner walls, at oesophageal end of spermathecal apparatus, representing 
the ampulla ; c, circular muscular libres of body-wall; oes., oesophagus; s.gl., septal glands; spth., 
main portion of spermathecal apparatus, representing the duct ; v., vacuole in surface epithelium. 



SOME LITTORAL OLIGOCH^ETA OF THE CLYDE. 65 

Fig. II. Sketch illustrating the appearance of the spermatheca? of Enchytrxus dubius in the living 
animal. 

Fig. 15. Anterior part of the body of Enchytrxus sabulosus; semi-diagrammatic, from the living animal. 
Amp., ampulla of spermatheca ; e.g., cerebral ganglion; d., duct of spermatheca; ass., oesophagus; 
pnph., peptonephridium ; ph., pharynx ; pr., prostomium ; s.g. 1 ' 3 , the septal glands of segments 
iv.-vi. ; i.—vii., segments i.-vii. 
Fig. 16. Anterior part of the body of Enchytrxics albidus, semi-diagrammatic. 

Amp., ampulla of spermatheca; e.g., cerebral ganglion; comm., communicating cord from septal 
glands to pharynx, splitting up into numerous smaller strands at its junction with pharynx ; 
d., duct of spermatheca; ws., oesophagus; pnph., peptonephridium; ph., pharynx; pr., 
prostomium; s., dark spot in cerebral ganglion; s.gl. 1 ~ s , the three pairs of septal glands. 
Fig. 17. Ventral portion of a transverse section through the penial bulbs of Fridericia hdbosa. x 250. 
A/,., alimentary canal, ventral to which are seen the ventral nerve-cord and ventral vessel; p., 
penial bulb of one side ; sp., septum \\ bulged backwards to form a sperm-sac ; <£ , male 
aperture. 
Figs. I, 6, 7, 8, 10, 11, 12, 13, 17 drawn from sections by means of Zeiss's Abbe's drawing apparatus. 



Trans. Roy. Soc. Edin r - 

Stephenson : Some Littoral Oligoch^eta of the Clyde — Plate I. 



Vol. XLVIII 



Mat 



mr 



Cffmyn 




Canvtn,'. 





>..- - 







....-** 



If 



# 



''St. ». 



<*s. 






1rn "n r "^ rl 'wnwmmmiiTrif^'" 



iWMiniwiiMmi» ii\m,«, 1 .\.i«luto Bin 




«»*&. 




Geo Waters to n,& Sons .Lilii . Edinburgh.. 



Trans. Roy. Soc. Edin r - y j XL VI II 

Stephenson: Some Littoral Oligocileta of the Clyde — Plate II. 




LJeo Watcreton&Sons.Luh. Edinburgh 



^ c 




( 67 ) 



Itr. — Les Mousses de l'Expedition nationale antarctique ecossaise. Par Jules 
Oardot. Presents par le Professeur I. Bayley Balfodr, M.D., F.R.S. (Avec 
trois Planches.) 

(MS. received March 9, 1911. Read June 19, 1911. Issued separately July 3, 1911.) 

AvERTISSEMENT. 

Les Mousses rapportees par l'Expedition nationale antarctique ecossaise ne sont pas 
tres nombreuses, mais elles presentent cependant de l'interet, parce que la plupart 
proviennent de localites qui etaient restees j usque la totalement inexplor^es. Le plus 
grand nombre des especes a ete recolte a l'ile Gough ou Diego Alvarez, dont la ilorule 
bryologique etait tout a fait inconnue ; les recoltes de M. le Dr R. N. Rudmose Brown 
nous ont fourni 21 especes pour cette petite ile. Dix especes ont ete recueillies a l'ile 
Laurie, l'une des Orcades meridionales, dans la region antarctique proprement dite ; 
enfin, 6 especes proviennent de 1'AscensioD. 

Nous etudierons separement les especes de ces trois localites. Des listes provisoires, 
mais incompletes et partiellement inexactes, ont ete pubises en 1905 et en 1906 par 
M. C. H. Wright, dans Linnean Society's Journal et dans Transactions and 
Proceedings of the Botanical Society of Edinburgh. 

J'ai a remercier MM. le Dr W. S. Bruce et le Dr R. N. Rudmose Brown, ainsi que 
les autorites du Jardin botanique de Kew, qui ont bien voulu mettre a ma disposition les 
materiaux de ce travail. 

Charleville, 
\ er decembre 1910. 

I.— MOUSSES DE LlLE LAURIE. * 

L'ile Laurie fait partie du groupe des Orcades meridionales, qui appartient deja 
au domaine antarctique proprement dit. M. Rudmose Brown y a recueilli 10 especes de 
Mousses, dont on trouvera plus loin l'enumeration ; mais il m'a communique en outre 
une serie de 6 especes recoltees en 1904, sur ia meme ile, par M. L. H. V alette, de 
l'Observatoire meteorologique de la Republique Argentine. Quatre des especes 
rapportees par M. Valette ne se trouvant pas dans les recoltes de M. Rudmose Brown 
le chiffre des Mousses actuellement constatees a l'ile Laurie se trouve ainsi porte a 
14, dont voici la liste : 

* Voir Cardot, "La Flore bryologique des Terres niagellaniques, de la Georgie du Slid et tie l'Antarctide," 
pp. 243-244 (IVissensch. Enjebn. der schwed. Sudpolar-Exped., 1901-1903, Bd. iv. Lief 8). 

TRANS. ROY. SOC. EDIN., VOL. XLVIIL, PART I. (NO. 3). 11 



68 



M. JULES CARDOT. 



Andrexa depressinerris Card. 
Dicranoweisia grimmiacea Broth. 
Dicranum aciphyllum Hook. fil. et Wils. 

„ Nordenskjoldii Card. 
Blindia Skottsbergii Card. 
Distichium capillaceum Br. et Sch. var. 

brevifolium Br. et Sch. 
Ceratodon purpureus (?) Brid. 



Grimmia An tar did Card. 

,, apocarpa Hedw. 
Weber a Racovitzx Card. 
Polytrichum alpinum L. 

,, subpiliferum Card. 

Brachythedum antartficum Card. var. 

cavi folium Card. 
Drepanodadus undnatus (Hedw.) Warnst. 



Sur ces 14 especes, aucune n'est speciale a l'lle Laurie, mais 4 : Andresea depressi- 
nervis, Grimmia Antarctici, Webera Racovitzse et Brachythedum antarcticum, 
n'ont pas ete rencon trees jusqu'ici en dehors du domaine antarctique. Quant aux 
autres especes, voici leur dispersion : 

Dicranoweida grimmiacea. Georgie du Sud, Kerguelen. 
Dicranum aciphyllum. Georgie du Sud, domaine magellanique 

,, Nordenskjoldii. Georgie du Sud. 
Blindia Skottsbergii. ' Georgie du Sud. 

Distichium capillaceum. Plus ou moins cosmopolite ; domaine magellanique. 
Ceratodon purpureus. Cosmopolite ; domaine magellanique. 
Grimmia apocarpa. Cosmopolite ; domaine magellanique. 
Polytrichum alpinum. Cosmopolite ; domaine magellanique, Georgie du Sud, Kerguelen. 

,, subpiliferum. Domaine magellanique. 

Drepanodadus undnatus. Cosmopolite ; domaine magellanique, Georgie du Sud, Kerguelen. 

Andre^ace^:. 

Andresea. 

A. depressmervis Card., in Rev. bryol., 1900, p. 43, et Result, voyage " Belgica.' 
Mousses, p. 22, pi. i., figs. 22-33. 

Andrexa sp. Wright, in Trans, and Proceed. Bot. Soc. Edinb., xxxiii., part i. 

Weisiace.-e. 
Dicranoive isia. 
D. grimmiacea (C. Mull.) Broth., in Nat. Pfianzenfam., Musci, p. 318. 



DlCRANACE^E. 

Dicranum. 

D. aciphyllum Hook. fil. et Wils., in Load. Journ. of Bot. , 1844, p. 541. 

D. Nordenskjoldii Card., in Bull. Herb. Boissier, 2 hme ser., vi. p. 14, et Fl. bryol. 
Terres magell., etc., pp. 265-266, fig. 59. 

Qampylopus introflexus Wright, loc. cit., non Mitt, 



LES MOUSSES DE L'EXPEDITION NATIONALE ANTAKCTIQUE ECOSSAISE. 69 

Seligeriace^. 

Blindia. 

B. Skottsbergii Card., in Bull. Herb. Boissier, 2 feme ser., vi. p. 4, et Fl. bryol. 
Terres magell., etc., p. 207, fig. 44. 

Campylopus vesticaulis Wright, loc. cit., non Mitt. 

Grimmiacejs. 
Grimmia. 
G. apocarpa (L.) Hedw., Sp. Muse, p. 76. 

" G. cf. apocarpa Hedw.," Wright, loc. cit. 

G. Antarctici Card., in Bull. Herb. Boissier, 2^ me ser., vi. p. 15, et Fl. bryol. 
Terres magell., etc., p. 271, pJ. v. figs. 16-25, pi. vi. figs. 1-5. 

G. amblyophylla Wright, loc. cit., non C. Mull. 

BRYACE.E. 

Webera. 

W. Racovitzse Card., in Rev. bryol., 1900, p. 44, et Result, voyage " Belgica," 
Mousses, p. 35, pi. xiii. figs. 1-14. 

Bryum sp. Wright, loc. cit. 

PoLYTRICHACEiE. 

Polytrichum. 

P. subpiliferum Card., in Rev. bryol., 1900, p. 42, et Result, voyage "Belgica,' 1 
Mousses, p. 39, pi. xii. figs. 1-14. 

Hypnacejs. 
Drepanocladus. 
D. uncinatus (Hedw.) Warnst., Beih. zum Bot. Centralbl., xiii. p. 417. 



70 M. JULES CARDOT. 

II.— MOUSSES DE L'lLE GOUGH OU DIEGO ALVAREZ. 

Sur les 21 especes de Mousses recoltees a l'ile Gough par M. Rudmose Brown, 
1 1 especes sont endemiques ; clu moms, elles ne me semblent pas pouvoir etre rap- 
portees a des especes signalees ailleurs. Un Dicranella, represente seulement par la 
plante male, est indeterminable. Restent 9 especes, sur lesquelles 6 appartiennent a la 
flore magellanique : 

Rhacomitrium symphyodontum Jaeg. Existe aussi au Chili, en Tasmanie et en Nouvelle-Zelande. 

Rhacomitrium subnigritum Par., represente" a l'ile Gongh par une variete endemiqne. 

Weber a nutans Hedw. ) ^ ■ ,., 

> rlus ou moms cosmopolites. 
,, albicans Sell. ) 

PoJytrichadelphus mayellanicus Mitt. Existe aussi dans la region australo-neozelandaise. 

Brachythecium subpilosum Jaeg. Se retrouve encore aux iles Marion, Kerguelen, Georgie du Sud et 

dans l'Antarctide. 

Deux especes se retrouvent a Tristan d'Acunlia : 

Rhacomitrium symphyodontum Jaeg. = R. membranaceum Par. 
Philonotis capillata Par. 

et une a 1' Ascension : 

Sphagnum Scotise Card. 

Enfin, une derniere espece : Cyclodictyon Isetevirens Mitt., existe en Irlande, a 
Madere et a Fernando- Po. 

Les especes endemiques montrent des affinites avec des Mousses de Tristan 
d'Acunha, de la region magellanique, de l'Afrique australe, et meme de la Reunion, de 
l'ile St Paul et de Kerguelen, dans l'Ocean Indien, mais c'est, en somme, avec la vegeta- 
tion de la region magellanique que la florule bryologique de l'ile Gough parait avoir le 
plus de rapports. 11 est toutefois probable que quand les Mousses de Tristan d'Acunha 
et celles de l'ile Gough seront mieux connues, on relevera un plus grand nombre 
d'especes communes a ces deux iles, qui presentent les plus grandes analogies quant a 
la flore superieure. 

Sphagnace^;. 

Sphagnum. 
S. Scotise Card. sp. nova. 

S. acutifolium Wright, in Linn. Soc. Journ., Bot., xxxvii. p. 264, non Ehrh. 

Molle, pallide viride. Caulis cellulee epidermicas distinctae, magnse, bistratosas, 
cylindrum lignosum pallidum, cellulis vix vel parum incrassatis formatum. Rami 3 
vel 4 in singulo fasciculo, quorum 1 vel 2 penduli. Folia caulina magna, 175-2 
millim. longa, 0'8— I millim. lata, oblongo-lingulata, basi haud vel vix angustata, apice 
obtuso, integro, plus minus cucullato, superne vel fere e basi fibrosa, limbo angusto 



LES MOUSSES DE L'EXPEDITION NATIONALE ANTARCTIQUE ECOSSAISE. 71 

ubique aequilato marginata. Folia ramorum clivergentium ovato-lanceolata, concava, 
1 "5—1 "6 millim. longa, 0*7-0 7 5 lata, marginibus superne inflexis, apice truncatulo et 
denticulato ; leucocystse valde fibrosse, poris majusculis, in parte superiore paginae 
dorsalis secundum chlorocystas sat numerosis, in pagina ventrali nullis vel perpaucis ; 
chlorocystse ventrales, in sectione transversali trapezoidales, utraque pagina inter 
leucocystas emergentes. 

Je n'ai vu que deux petits fragments de cette espece, Tun provenant de l'ile Gough, 
l'autre de 1' Ascension. Elle est voisine du S. Reichardtii Hpe., de l'ile St Paul, mais 
celui-ci a les feuilles caulinaires plus courtes, ovales et a leucocystes toutes divisees par 
plusieurs cloisons obliques, ce qui n'a lieu, dans l'espece nouvelle, que sur un petit 
nombre de leucocystes. 

DlCRANACE/E. 

Trematodon. 

T. intermixtus Card. sp. nova. 

Aliis muscis commixtus gregarie crescens. Caulis gracilis, mollis, erectus, laxifolius, 
6-10 millim. longus. Folia mollia, e basi subvaginante breviter oblonga in subulam 
elongatam, canaliculatam, plus minus flexuosam, integerrimam vel apice minute 
denticulatam sat abrupte constricta, media et superiora 4'5-5'5 millim. longa, 0'6-0'75 
basi lata, inferiora breviora, costa basi angusta, superne dilatata et totam fere subulam 
occupante, cellulis basis elongatis, linearibus, in subula brevioribus, minute rectangulis. 
Folia perichgetialia longiora, e basi laxius reticulata magis sensim angustata. Capsula 
in pedicello pallide stramineo, 12-15 millim. longo erecta inclinatave, setate arcuata, 
collo sporangio longiore basi strumuloso instructa, 3-4 millim. longa, operculo 
longirostro. Peristomii dentes anguste lanceolati, circa 0'35 millim. longi, rubro- 
aurantiaci, dorso longitudinaliter striati, intus papillosi, lamellis paucis ornati, usque ad 
basin in 2 crura apice coliasrentia divisi. Sporse luteo-virides, minute granulosa?, diam. 
18-20 m. Flores masculi gemmiformes, aggregati, terminales. 

Se rapprochant par ses feuilles longuement subulees du T. sctaceus Hpe., de l'ile 
St Paul, cette espece m'en parait suffisamment distincte par sa capsule a col plus long 
que le sporange, et par ses dents peristomiales divisees jusqu'a la base en deux branches 
distinctes, plus ou moins coherentes seulement au sommet. Les echantillons trop 
pauvres dont je disposals ne m'ont pas permis de reconnaitre si les fleurs males naissent 
sur des tiges speciales, ou bien au sommet de rameaux basilaires de la plante fructifere. 

Dicranella. 

D. sp., planta mascula. 

Probablement une espece nouvelle, dont nous n'avons malheureusement que la 
plante male. Petite Mousse de 2a 4 millimetres, a feuilles etalees-dressees, flexueuses, 
planes aux bords, a subule generalement plus ou moins obtuse ou un peu tronquee et 
denticulee au sommet. 



72 M. JULES CARDOT. 

Campylopus. 

C, alvarezianus Card. sp. nova. 

Cespites superne lutescentes, intus fusco-tomentosi, 1-4 centim. alti. Caulis 
simplex vel parce divisus, ssepe basi raraos filiformes gracillimos emittens. Folia plus 
minus conferta, superiora comosa, subsecunda, anguste lanceolata et sensim in subulam 
canaliculatam, acutam, dorso scaberulam, apicem versus dentatam, rarias subintegram 
protracta, 4 '5-5 millim. longa, - 5-0"65 millim. basi lata, inferiora minora, appressa, costa 
latissima, |— § basis et totam fere subulam occupante, elamellosa, in sectione transversali 
a cellulis ventralibus majusculis, eurycystis dorso stereidis et substereidis tectis, 
cellulisque epidermicis dorsalibus composita, cellulis alaribus tenerrimis, hyalinis, 
marcescentibus, parum distinctis, cseteris lineari-rectangulis et subquadratis, parietibus 
incrassatis. Reliqua desiderantur. 

On peut comparer cette espece au C. vesticaulis Mitt., de Tristan d'Acunha, mais celui- 
ci est plus robuste, ses tiges sont recouvertes d'un tomentum plus abondant, et ses feuilles, 
plus grandes, presentent dans la partie moyenne un tissu fort different, compose de 
cellules irregulieres, plus ou moins obliques, attenuees, subrhomboidales. Le C. alvarezi- 
anus rappelle assez, par son aspect exterieur, le C. eximius Reich., de l'ile St Paul, mais 
s'en separe d'ailleurs completement par ses feuilles epiliferes et par son tissu. 

J'ai trouve dans les recoltes de M. Rudmose Brown quelques tiges d'un Campy- 
lopus a feuilles plus molles, plus flexueuses a l'etat sec, et a tissu forme j usque pres de 
la basede cellules plus courtes, carrees ou brievement rectangulaires, qui, bien qu'assez 
different de l'espece que je viens de decrire, me semble cependant n'en etre qu'une 
simple forme. 

Grimmiacejs. 

Khacomitriuni. 

R. symphyodontum (C. Miill.) Jaeg., Ad., i. p. 375. 

R. Jiavescens Card., in Rev. bryol., 1900, p. 41 ; Wright, loc. cit., pro parte. 

Echantillon sterile. 

La forme recoltee a l'ile Gough par M. Rudmose Brown ne ditfere pas de la plante 
magellanique, qui est tres variable ; mais elle ne represente pas exactement le 
R. Jiavescens Card., que jc ne considere plus, d'ailleurs, que comme une des nombreuses 
formes du R. symphyodontum. Le R. membranaceum (Mitt.) Par., de Tristan 
d'Acunha, ne me parait etre egjalement qu'une forme de la meme espece, caracterisee par 
ses feuilles etroites et son pedicelle extremement court. 

R. subnigritum (C. Miill.) Par., Ind. bryol., ed. i., p. 1080. Var. alvarezianum 
Card, var nova. 

R. Jiavescens Wright, loc. cit., pro parte. 

A forma typica patagonica et fuegiana differt : colore minus nigricante, obscure vel 



LES MOUSSES DE L'EXPEDITION RATIONALE ANTA RCTIQTJE ECOSSAISE. 73 

sordide viridi, foliis majoribus, latioribus (3"3-375 millim. longis, 1 — 1*2 latis), 
mollioribus, siccitate minus imbricatis, margin ibus minus late incrassatis, costaque 
validiore. basi 180-220 m lata (loco 112-140 in forma genuina). Sterile. 

Orthotrjohace^e. 

Macromitrium. 

M. antarcticum Wright, in Linn. Soc. Journ., Bot., xxxvii. p. 264. 

Cespites densi, lutescenti-virides. Caulis repens, ramis confertis, brevissimis, 
subnodulosis dense pinnatus. Folia conferta, sicca cirrata, madida patenti-erecta, anguste 
oblongo-lanceolata vel subligulata, carinata, acuminata, acuta obtusulave, integerrima, 
1-T8 millim. longa, 0'25-0"45 lata, marginibus ubique planis vel inferne anguste 
reflexis, costa percurrente vel subpercurrente, cellulis omnibus lsevissimis, inferioribus 
vermicularibus, angustissimis, parietibus perincrassatis, caeteris quadratis vel sub- 
rotundatis. Folia perichaetialia intima caulinis latiora, oblongo-lanceolata. Capsula 
in pedicello lsevi, 4-6 millim. longo, erecta, ovata, pachyderma, 1-1 "25 millim. longa, 
0"5-0 - 7 lata, ore rubro, vernicoso, siccitate plicato, operculo longirostri. Peristomium 
simplex, dentibus griseis, granulosis, truncatis. Calyptra nuda. 

Espece de la section Goniostoma, tres voisine du M. borbonicwn (Besch.) Broth., 
mais ayant les feuilles plus longues, les capsules et les pedicelles plus courts ; elle se 
rapproche aussi beaucoup du M. Seemanni Mitt., de Ste Helene, qui s'en distingue par 
son port plus robuste, sa teinte d'un jaune brun, ses feuilles plus retrecies dans le haut, 
les cellules allongees occupant une plus grande partie de la feuille et s'avancant jusqu'au 
dela du milieu (tandis qu'elles s'arretent generalement au dessous du milieu dans le 
M. antarcticum), les cellules superieures plus arrondies, a parois plus epaissies, 
jaunatres, le pedicelle plus epais, et la capsule plus solide. 

Bryace^e. 

Webera. 

W. nutans (Schreb.) Hedw., Sp. Muse, p. 168. 

Quelques tiges depourvues de capsules, melangees au Canipylopus alvarezianus ; 
inflorescence paroique ou subsynoique. Parait bien identique au type de l'hemisphere 
bor6al. 

W. albicans (Wahlenb.) Sch., Coroll., p. 67. 

Tiges st^riles, au milieu des gazons de Philonotis capillata. C'est une forme grele, 
comme on en trouve egalement en Europe. 

Br yum. 
B. tenellicaule Card. sp. nova. • 

Cespites tenelli, densiusculi, nitiduli, viridi-lutescentes, laxe cohserentes. Caulis 
gracillimus, ruber, laxifolius, parce radiculosus, 7-12 millim. altus, simplex vel parcissime 



74 M. JULES CARDOT. 

divisus. Folia siccitate paten ti-erecta, subflexuosa, madore patentia, caviuscula, anguste 
lanceolata, acuminata, costa excurrente cuspidata, 1 "25— 1*5 millim. longa, 0*25-0 "45 lata, 
marginibus nunc planis, nunc reflexis vel anguste revolutis, apicem versus remote et 
minute denticulatis, costa valida, basi 50-70 m lata, viridi vel lutescente, in subulam 
crassam, parce denticulatam vel subintegram longiuscule excedente, cellulis inferioribus 
rectangulis et subrectangulis, caeteris oblongo-rhomboideis. Flores fructusque 
desiderantur. 

Cette petite espece, de la section Doliolidium, pent etre comparee au B. coronatum 
Schw. ; elle en differe par sa petite taille, ses tiges plus greles, ses feuilles beaucoup plus 
petites, etc. 

B. subidinerve Card. sp. nova. 

Cespites densiusculi, pallide vel sordide virides. Caulis superne dense, inferne 
laxius foliosus, 6-12 millim. altus, dichotome divisus et subfastigiato-ramosus. Folia 
madida patentia, sicca suberecta, concava, inferiora lanceolata, acuminata, superiora 
late ovato-lanceolata, brevius acuminata, costa longe excurrente cuspidata, 1*3-1 '6 
millim. longa, 0*5-0*7 lata, marginibus plerumque e basi usque apicem versus revolutis, 
rarius subplanis, integris vel superne sinuato-subdenticulatis, costa valida, 70-80 fi 
basi lata, viridi-lutescente, in subulam crassam, remote denticulatam longe excedente, 
cellulis mediis et superioribus oblongo-rhomboideis, parietibus crassiusculis, inferiori- 
bus breviter rectangulis et subquadratis, infimis laxis, teneris, rubellis vel subhyalinis. 
Csetera desiderantur. 

Appartenant egalement a la section Doliolidium, cette espece se distingue du B. 
coronatum Schw. par ses feuilles generalement revolutees, pourvues d'une nervure plus 
forte, formant une subule plus epaisse et plus longue, les feuilles superieures plus 
larges et plus courtes. 

Bartramiace^e. 

Bartramia. 

B. stenobasis Card. sp. nova. 

Cespites densi, lutescenti-virides. Caulis erectus, simplex, parum radiculosus, 1*5-2 
centim. altus. Folia sicca et madida erecto-rlexuosa vel patenti-erecta, fragilia, 
facillime decidua, e basi parva, angusta, vix dilatata longissime subulata, setacea, 
utraque p;igina papillosa, 4-5 millim. longa, basi vix 0*12 lata, marginibus serrulatis, 
costa dilatata, in subulam dentatam, scabram exeunte, cellulis basilaribus laxis, pellucidis, 
elongatis, lsevibus, caeteris linearibus, angustis, 2-3-stratosis, parietibus trans versis 
prominentibus papillosis. Caetera desunt. 

Rappelle assez le B. patens Brid., mais en differe par ses feuilles a partie basilaire 
plus petite, plus etroite et moins brusquement contractee. Espece remarquable par 
le peu dc developpement de la partie basilaire de la feuille, tres differente du B. 
radicosa Mitt, de Tristan d'Acunha, qui est beaucoup plus robustc, et a les feuilles 



LES MOUSSES DE L'EXPEDITION NATIONALS ANTARCTIQUE ECOSSAISE. 75 

moins finement subulees, brusquement et fortement dilatees a la base, et les tiges tres 
radiculeuses. 

Philonotis. 

Ph. capillata (Mitt.) Par., lad. bryol., ed. i., p. 919. 

Echantillons steriles et plante male. 

II y a deux formes differentes dans les recoltes de M. Rudmose Brown. L'une est 
completement identique a la plante originale de Tristan d'Acunha ; l'autre est plus grele, 
plus petite, d'une vert glauque, plus molle dans toutes ses parties ; mais elle ne differe 
pas autrement du type. Cette derniere forme croissait intimement melangee au Webera 
albicans, dont elle a un peu l'aspect. 

PoLYTRICHACE^E. 

Polytrichadelphus. 
P. magellanicus (L.) Mitt., in Journ. Linn. Soc, 1859, p. 97. 
Polytrichum commune Wright, in Linn. Sue. Journ., Bot., xxxvii. p. 265, non Linn. 

Tiges steriles, mais la structure de la feuille et des lamelles ne laisse aucun doute 
sur leur determination. 

HOOKERIACE^E. 

Oyclodictyon. 

C. Isetevirens (Hook, et Tayl.) Mitt., in Journ. Linn. Soc, 1864, p. 163. 
Echantillon sterile, bien identique a ceux d'Irlande. 

Leskeace^e. 
Thuidium. 

Th. citvarezianum Card. sp. nova. 

Humile, graeile. Caulis primarius repens, tenellus, secundarius erectus ascen- 
densve, 1-2 centim. longus, remote et irregulariter pinnatus et parcissime bipinnatus, 
paraphylliis sat numerosis, brevibus, simplicibus, papilloso-dentatis obtectus. Folia 
madida undique patentia, sicca incurvato-crispata, caulina e basi late cordata abrupte 
acuminata, 0"4-0 '5 millim. longa, '2 5-0 '3 5 lata, ramea aliquid minora, magis sensim 
latiuscule acuminata, 0*3-0 '4 millim. longa, 0*1 8-0 "20 lata, ramulina minima, ovato- 
lanceolata, 0*15-0*18 millim. longa, vix 0*08 lata, omnia caviuscula, acuta, marginibus 
planis, crenulatis. superne serrulatis, costa in acumine evanida, cellulis quadratis vel 
subrotundatis, utraque pagina papilla singula medio notatis. Csetera desiderantur. 

Cette espece se rapproche du Th. curvatum Mitt., de Tristan d'Acunha; elle en 
differe par sa taille plus faible, son port beaucoup plus grele, et ses feuilles caulinaires 

TRANS. ROY. SOC. EDIN., VOL. XLVIIL, PART I. (NO. 3). 12 



76 M. JULES CARDOT. 

et rameales moins dimorphes, plus petites, plus courtes et plus brievement acuminees. 
Le Th. curvatum, que M. Brotherus place dans les Thuidiella, est certainement, 
d'apres 1'echantillon original que j'ai pu examiner, un Thuidiopsis, tres voisin des 
Th. unguiculatum (Hook. fil. et Wils.), furfurosum (Hook. fil. et Wils.) et hastatum 
(C. Miill.), de la Nouvelle-Zelande. C'est done egalement dans la section Thuidiopsis 
que doit prendre place l'espece nouvelle. 

Hypnace^. 

Isopterygium. 

I. Brownii Card. sp. nova. 

Tenellum, intricato-repens, lutescenti-viride, nitidulura. Caulis gracillimus, 8-12 
millim. longus, irregulariter ramosus, ramis complanatulis, attenuatis. Folia laxiuscula, 
subdistiche patentia vel sursum leniter homomalla, anguste lanceolata, sensim longeque 
acuminata, lateralia falcatula, obsolete binervia vel enervia, media 1 '1-1 '35 millim. 
longa, 0'25-0'37 lata, marginibus planis, superne serrulatis, cellulis anguste linearibus, 
mediis longissimis, alaribus perpaucis subindistinctis. Csetera desiderantur. 

Cette petite espece rappelle assez les I. antarcticum (Mitt.) Card, et fuegianum 
Besch. ; elle sen distingue par ses feuilles etroitement lanceolees et terminees par un 
acumen moins long, moins etroit et denticule. Elle croissait au milieu des tiges du 
Bartramia stenobasis. 

A propos de VI. antarcticum, je ferai remarquer que la Mousse de Kerguelen que 
C. Muller a decrite en 1890 sous le nom de Hypnum (Plagiothecium) antarcticum 
[Forschungsreise S.M.S. " Gazelle," Laubmoose, p. 34) n'est nullement le 
Plagiothecium antarcticum de Mitten, qui est un Isopterygium, tandis que la 
plante de Muller est un Plagiothecium. Muller reconnaissait, d'ailleurs, qu'il n'etait 
pas certain de l'identite des deux plantes, qui, de fait, sont fort differentes. Mais le 
P. antarcticum C. Miill. non Mitt., de Kerguelen, est exactement la meme chose 
que l'espece decrite l'annee precedente par Muller sous le nom de Hypnum 
[Plagiothecium) georgico-antarcticum ('"' Bryologia Austro-Georgiae," in Ergebn. der 
deutsch. Polar-Exped., All. Theil, Bd. ii., 11, p. 321). Les differences que l'auteur 
indique entre les deux plantes ne sont pas constantes et n'ont aucune valeur : le tissu 
des feuilles n'est pas plus chlorophylleux dans Tune que dans l'autre, et l'acumen est 
souvent denticule au sommet sur la Mousse de la Georgie du Sud. C'est l'espece de 
Muller, et non celle de Mitten, que M. Brotherus a mentionnee sous le nom de 
Plagiothecium antarcticum dans son tableau synoptique du genre (in Engler et 
Prantl, Pjianzenfamil., Musci, p. 1086). J'ajouterai que le Hypnum austropulchellum 
de Muller [Forschungsreise, etc., p. 35) pourrait bien etre l'espece de Mitten. 

/. ambiguum Card. sp. nova. 

Molle, lutescens, nitidulum, robustulum, intricato-cespitosum. Caulis 2-3 centim. 
longus, irregulariter divisus, ramis flaccidis, complanatis, obtusis. Folia compressa, 



LES MOUSSES de l'expedition nationale antarctique ecossaise. 77 

distiche patentia vel subhomomalla, e basi ssepe subdecurrente oblongo-lanceolata, 
longiuscule et acute acuminata, lateralia aliquid asymmetrica et curvatula, 2-2*5 millim. 
longa, '5-0 '75 lata, marginibus planis ubique integris vel apicem versus remote et 
minute denticulatis, costa gemella vel furcata, ad £— J producta obsoletave, cellulis 
anguste linearibus, flexuosis, mediis longissimis, alaribus plerumque distinctis, laxis, 
ovatis, oblongisve, subinflatis. Csetera ignota. 

En raison de ses cellules alaires ordinairement assez diiferenci^es et souvent sub- 
ddcurrentes, cette espece occupe une place indecise eDtre les genres Isopterygium et 
Plagiothecium. 

Brachytkecium. 

B. pallidofiavens Card. sp. nova. 

Gracile, pallidofiavens, nitidulum. Caulis longe repens, flexuosus, rhizoidis 
fasciculatis radiculosus, irregulariter pinnatus, ramis teretibus, patulis, siccitate 
julaceis, breviter attenuatis. Folia ramea madida erecto-patentia, sicca erecta, 
subappressa, oblongo-lanceolata, sensim tenuiterque acuminato-subulata, plicata, 
1*5—1 "8 millim. longa, 0'35-0'5 lata, marginibus planis vel plus minus revolutis, inferne 
integris, superne remote et minute denticulatis, costa tenui, ad § evanida, reti pallido, 
cellulis anguste linearibus, parietibus crassiusculis, alaribus distinctis, quadratis vel 
rectangulis. Folia caulina laxiora, haud vel vix plicata, costa breviore, medium versus 
evanida. Csetera desunt. 

Cette espece, dont je n'ai trouv6 que quelques tiges en melange avec les autres 
Mousses, est voisine des B. austrosalebrosum et austroglareosum (C. Mull.) Par. ; elle 
differe du premier par ses dimensions plus faibles, par ses rameaux julac^s a l'6tat sec, 
et par ses feuilles plus, etroites, denticulees dans la partie superieure ; elle se distingue 
du second par ses rameaux plus greles, et par ses feuilles plus etroites, a bords plans 
ou moins regulierement revolutds. 



o 



B. subpilosum (Hook. fil. et Wils.) Jaeg., Ad., ii. p. 410. 

Un petit echantillon sterile, dont l'attribution a cette espece ne me parait cependant 
pas douteuse. 

Rhynchostegium. 

R. isopterygioides Card. sp. nova. 

R. rhaphidorhynchum Wright, in Linn. Soc. Journ., Bot., xxxvii. p. 265, non Hypnum raphidorrhynchum 
C. Mull., Syn., ii. p. 354. 

Autoicum, lutescens, nitidum. Caulis procumbens, vage pinnatus, ramis com- 
planatulis, isopterygioideis, obtusis. Folia caulina erecto-patentia, ramea compressula, 
late ovato-lanceolata, acuminata, acumine acuto plerumque semitorto, 1 '5-1 '75 millim. 
longa, 0'7-0'85 lata, marginibus planis e basi serrulatis, costa tenui, ad f evanida, 
cellulis pellucidis, linearibus, subflexuosis, alaribus paucis, brevioribus, subrectangulis 



78 



M. JULES CARDOT. 



et subquadratis. Folia perichsetialia intima e basi oblonga, convoluta, in acumen 
longiusculum serrulatum protracta. Pedicellus rubellus, lsevis, 10-12 millim. longus. 
Csetera ignota. 

Cette Mousse differe du R. raphidorrhynchum (C. Mull.) Jaeg., de l'Afrique 
australe, par ses feuilles plus fortement dentees, terminees par un acumen moins etroit, 
en general a demi tordu, et par son tissu moins serre. Elle se rapproche beaucoup du 
R. confertum Br. eur., d'Europe, mais s'en distingue cependant par son port, ses rameaux 
coinprimes, qui lui donnent l'aspect d'un I so pterygium, et ses feuilles plus dentees. 
Peut-etre est-ce la merae plante qui a 6te indiqu6e par Mitten a Tristan d'Acunha sous 
le nom de Hypnum raphidorrhynchum. 



Ill— MOUSSES DE L'ASCENSION. 

D'apres le Bryologia atlantica, oeuvre posthume du regrette A. Geheeb, qui vient 
de paraitre tout recemment, la fiorule bryologique de l'Ascension comprend 20 
especes ; ce chiffre se trouve maintenant porte a 24 par les recoltes de M. Rudmose 
Brown. Voici Enumeration complete de ces especes : 



Sphagnum Scotix Card. 
Dicranella pygmxa Card. 

,, ascensionica Mitt. 

Gampylopus smaragdinus (Brid.) Jaeg. 

,, introflexus (Hedw.) Mitt. 

,, Naumanni (C. Mull.) Par. 

Calymperes Ascensionis C. Mull. 
Gymnostomum Lessonii Besch. 

„ Bescherellei Broth, et Geh. 

Hyophila Ascensionis Card. 
Barbula leucochlora C. Mull. 
,, cuspidatissima C. Mull. 



Bryum zygodontoides C. Mull. 
,, argentatum C. Mull. 
,, rubrocostatum C. Mull. 
Philonotis penicillata Wright. 
,, per gracilis Card. 
,, subolescens (C. Mull.) Par. 
Leucodon Bescherellei Broth, et Geh. 
Neckera Ascensionis Besch. 
Callicostella Ascensionis C. Mull. 
Rhacopilum gracile Mitt. 

,, Naumanni C. Mull. 

Taxit helium planum Brid. 



II est fort possible que le Rhacopilum gracile Mitt. (1885) soit la meme espece que 
le R. Naumanni C. Mull. (1883). Le Gymnostomum Bescherellei est une espece 
nouvelle, qui est figuree dans l'ouvrage de Geheeb ; le Leucodon Bescherellei est une 
autre espece nouvelle, malbeureusement restee a l'etat de nomen nudum. 

Sauf trois, toutes les especes sont speciales a l'ile de l'Ascension. Les trois especes 
non endemiques sont : 

^ Sphagnum Scotix Card., qui se retrouve, ainsi que nous l'avons vu, a l'ile Gough. 

Gampylopus introflexus (Hedw.) Mitt. (C. polytriclioides De Not.), plus ou moins cosmopolite. 
J "axil 'helium planum Brid., esp&ce de l'Amerique tropicale, dont l'existence r^elle a l'Ascension reste 
bien douteuse. 



LES MOUSSES DE L'EXPEDITION RATIONALE ANTARCTIQUE ECOSSAISE. 79 

Sphagnace^e. 
Sphagnum. 
S. Scotise Card, (vide supra, p. 70). 
S. cuspidatum Wright, in Trans, and Proceed. Bot. Soc. Edinb., xxiii., ii. p. 203, non Ehrh. 

Le petit fragment que j'ai vu provenant de 1' Ascension ne parait pas differer de la 
Sphaigne de l'ile Gough. 

DlCRANACE^E. 

Dicranella. 

D. pygmsea Card. sp. nova. 

Dioica, humillima, lutescenti-viridis, 5-6 millim. alta. Folia erecta vel leniter 
subsecunda, anguste triangulari-lanceolata, sensim in acumen canaliculatum, crassius- 
culum, integrum, acutum vel obtusulum producta, 0"9-l'35 millim. longa, 0'18-0*25 
lata, marginibus superne inflexis, cseterum planis et ubique integerrimis, costa valida, 
lutescente, bene limitata, quartam vel tertiam partem basis occupante, continua vel 
subexcurrente, cellulis oblongis, rectangulis et linearibus, parietibus firmis, incrassatis. 
Folia perichsetialia majora, anguste oblongo-lanceolata, laxius reticulata. Capsula in 
pedicello pallido, circa 2 millim. longo, siccitate apice leniter dextrorsum torto minima, 
erecta vel suberecta, sicca ovata, madida subglobosa, aperta late truncata, circa 0*5 
millim. longa, 0*3-0 "4 lata, operculo longirostri capsulse sequilongo. Annulus duplex 
et triplex. Peristomium rudimentarium, dentibus minimis, rubellis, irregularibus, 
annulo vix sequilongis et quidem brevioribus. Planta mascula ignota. 

Tres voisine du D. minuta (Hpe.) Broth., de Madagascar, cette espece en differe 
cependant par ses feuilles plus longues et plus etroitement acumin<5es. Elle etait 
melangee a l'espece suivante. 

D. ascensionica Mitt., in Mellis, St Helena, p. 357. 

Par son pedicelle fortement flexueux et courbe, cette espece se rapproche des 
Campylopodium, mais les feuilles sont moins brusquement dilat^es a la base que celles 
de ce genre. M. Bkotherus a fait d'ailleurs observer avec raison que le genre 
Campylopodium est tres faiblement caracterise, et qu'il serait peut-etre preferable de le 
considerer comme un sous-genre de Dicranella (Musci, in Pjianzenfamil., p. 312). 

Campylopus. 

C. smaragdinus (Brid.) Jaeg., Ad., i. p. 136. 

II y a, dans l'herbier du Museum de Paris, deux echantillons de cette espece. 
L'un, provenant de l'herbier Thuret, recolte par Lesson en 1829, et etiquete par 
Bescherelle, est du C. smaragdinus pur. L'autre, provenant de l'herbier Brongniart, 
estun ^chantillon de la plante originale recoltee par Dumont d'Urville en 1825, et sur 



80 M. JULES CARDOT. 

laquelle Bridel a etabli son Didymodon smaragdinus [Bryol. univ., i. p. 819). Cet 
echantillon 6tait etiquete primitivement " Thisanomitrium intrqflexum" puis a 
ete rapporte plus tard par Bescherelle au C smaragdinus. Mais, en reality, il com- 
prend deux especes : la plus grande partie de la touffe est bien du C. smaragdinus, au 
milieu duquel on trouve des brins d'une espece a feuilles piliferes, a nervure fortement 
lamellifere sur le dos, qui est, conformement a la premiere etiquette, du C. introjiexus 
(Hedw.) Mitt, si toutefois, avec Mitten, on reunit au Dicranum intrqflexum d'HEnwiG 
le C. polytrichoides De Not., mais a laquelle il faudrait attribuer ce dernier nom, si Ton 
reserve celui de C. introjiexus aux formes australes a poil refleclii ou plus ou moins 
^tale. 

L'echantillon des r6coltes de M. Rudmose Brown qui m'a 6t6 communique par 
l'Herbier royal de Kew, appartient au C. smaragdinus ; mais il est fort possible qu'il y 
avait dans la recolte de M. Brown le meme melange que dans celle de Dumont d'Urville, 
car sur la liste qui a et6 publiee dans Trans, and Proceed, of the Bot. Soc. Edin., vol. 
xxiii., part ii., on ne trouve cite que C. introjiexus, bien que les deux especes soient 
totalement differentes. 

Pottiace^e. 
Hyophila. 
H. Ascensionis Card. sp. nova. 
" Barbula cf. leucocldora C. Mull.," Wright, loc. eit. 

Cespites fusco-virides. Caulis erectus, apicem versus dichotome vel fastigiato- 
ramosus, 12-15 millim. altus. Folia siccitate crispata, madore erecto-patentia, oblongo- 
lingulata, brevissime acuminata vel subapiculata, in singula innovatione annua 
ascendendo majora, media et superiora 175-2"25 millim. longa, 05-07 lata, marginibus 
plus minus inflexis, superne irregulariter crenato-subdenticulatis, costa rufa, valida, 
100-120 m basi crassa, continua vel brevissime excedente, cellulis majusculis, subrotun- 
datis vel subquadratis, papillosis, chlorophyllosis, parietibus lutescentibus incrassatis, 
inferioribus rectangulis, pellucidis. Castera desiderantur. 

Cette espece rappelle assez, par son aspect general, le H. crenulatula C. Miill., du 
Cameroun, mais s'en distingue aisement par ses feuilles plus courtes, formes de cellules 
3 a 4 fois plus grandes. 

Bartramiace^e. 
Philonotis. 
Ph. pergvacilis Card. sp. nova. 
" Bariramia cf. suholescens C. Miill.," Wright, loc. cit. 

Cespites tenelli, virides, intus dense fusco-tomentosi. Caulis erectus, gracillimus, 
parcissime ramosus vel subsimplex, 15-25 millim. altus. Folia erecto-patentia, 



LES MOUSSES DE {/EXPEDITION NATION ALE ANTARCTIQUE ECOSSAISE. 81 

anguste lanceolata, sensim cuspidata, minima, 0*9-1 *1 millim. longa, 0*15-0'2 lata, 
marginibus plerumque e basi longe et anguste revolutis, apicem versus planis, ubique 
simpliciter serrulatis, costa basi 30-40 m crassa, dorso scabra, in cuspidem denticu- 
latam, validiusculam excedente, cellulis angustis, linearibus, parietibus transversis 
prominentibus, inferioribus laxioribus, rectangulis quadratisve. Caetera ignota. 

Bien distinct du Ph. subolescens (C. Mull.) Par. par ses tiges plus elancees, ses 
feuilles beaucoup plus longues, revolutees aux bords, et son tissu plus serre et plus 
chlorophylleux. Je ne connais pas le Ph. penicillata Wright, qui est egalement 
particulier a l'Ascension, mais il est probable que ce n'est pas la meme chose que la 
Mousse que je viens de decrire, puisque M. Wright, qui a vu celle-ci, n'y a pas reconnu 
son espece, et Fa rapproch^e de preference du Ph. subolescens. 



EXPLICATION DES PLANCHES. 



Planchb I. 



Fig. 1. Sphagnum Scotix. — a, feuille caulinaire ; x 13. b, c, feuilles d'un rameau divergent; x 13. 
d, tissu dans le haut d'une feuille caulinaire ; x 270. e, tissu dans la moitie superieure d'une feuille 
rameale, vu par la face dorsale ; x 270. /, portion d'une section trausversale vers le milieu d'une feuille 
rarneale ; x 270. 

Fig. 2. Trematodon intermixtus. — a, plantes, gr. nat. b, c, feuilles; x 13. d, e, capsules deoperculees ; 
x 13. /, fragment du peristome et spores; x 138. 

Fig. 3. Campylopus alvarezianus. — a, plante, gr. nat. b, c, d, feuilles; x 13. e, tissu basilaire d'une 
feuille; x 138. /, tissu vers le milieu d'une feuille ; x 270. g, sommet d'une feuille ; x 138. h, partie 
d'une coupe transversale de la nervure, dans la moitie superieure ; x 270. 

Fig. 4. Macromitrium antarcticum. — a, plante, gr. nat. b, c, d, e, feuilles ; x 26. /, tissu basilaire 
d'une feuille ; x 270. g, tissu vers le milieu d'une feuille ; x 270. h, sommet d'une feuille ; x 270. i, 
capsule jeune et encore operculee ; x 13. j, capsule mure, deoperculee, a l'etat sec; x 13. k, fragment du 
peristome; x 138. I, coiffe ; x 13. 

Fig. 5. Bryum tenellicaule.—a, plante, gr. nat. b, extremite d'une tige ; x 13. c, d, e, feuilles; x 26. 
/, tissu basilaire d'une feuille ; x 138. g, sommet d'une feuille ; x 138. 

Fig. 6. Bryum subulinerve.—a, b, plantes, gr. nat. c, extremite d'une tige; x 13. d, e, J\ g, feuilles; 
x 26. h, tissu basilaire d'une feuille; x 138. i, sommet d'une feuille; x 138. 

Planche II. 

Fig. 7. Bartramia stenobasis.—a, plante, gr. nat.; b, c, feuilles; x 13. d, tissu de la partie superieure 
de la base d'une feuille; x 138. e, tissu marginal vers le milieu d'une feuille; x 138. /, sommet d'une 
feuille; x 138. 

Fig. 8. Thuidium alvarezianum.—a, b, plantes, gr. nat. c, extremite d'une tige ; x ] 3. d, e, f, feuilles 
caulinaires ; x 32. g, h, i, feuilles d'un rameau primaire ; x 32. j, Jc, I, feuilles d'un rameau secondaire ; 
x 32. m, tissu marginal vers le milieu d'une feuille caulinaire ; x 270. n, sommet d'une feuille caulinaire ; 
x 270. o, paraphylles ;_ x 270. 

Fig 9. Isopterygium Brownii.—a, b, c, plantes, gr. nat. d, extremite d'une tige; x 13. e, /, g, h, 
feuilles ; x 26. i, tissu basilaire d'une feuille ; x 270. j, sommet d'une feuille ; x 270. 

Fig. 10. Isopterygium ambiguum.—a, plante, gr. nat. b, extremite d'une tige; x 13. c, d, e, f, g, 
feuilles ; x 13. h, tissu basilaire d'une feuille ; x 270. i, sommet d'une feuille ; x 270. 



82 LES MOUSSES DE L'EXPEDITION NATIONALE ANTARCTIQUE ECOSSAISE. 

Fig. 11. Brachythecium pallidoflavens. — a, plante, gr. nat. b, extremite d'un rameau ; x 13. e, d, e, 
feuilles ; x 26. /, tissu basilaire d'une feuille; x 270. g, tissu marginal dans la moitie superieure d'une 
feuille ; x 270. h, somniet d'une feuille ; x 270. 

Planche III. 

Fig. 12. Rhynchostegium isopterygioides. — a, plante, gr. nat. b, extremite d'un rameau; x 13. 
<*> d> e,f, g, feuilles ; x 13. h, tissu basilaire d'une feuille; x 138. i, tissu marginal vers le milieu d'une 
feuille; x 138. ./, sonimet d'une feuille ; x 138. k, feuille perichetiale intime ; x 13. 

Fig. 13. Dicranella pygmxa. — a, b, plantes ; x 3. c, d, e, feuilles; x 26. /, tissu basilaire d'une 
feuille; x 138. g, sommet d'une feuille; x 138. k, feuille pericbetiale ; x 26. i, j, capsules operculees, 
a l'etat sec ; x 26. k, capsule mure, ouverte, a l'etat humide ; x 26. Z, fragment du peristome et de 
l'anneau ; x 138. m, feuille de D. minuta (Hpe.) Broth. ; x 26. 

Fig. 14. Hyophila Ascensionis. — a, b, plantes, gr. nat. c, extremite d'une tige ; x 13. d, e, f, g, h, 
feuilles; x 13. i, tissu basilaire d'une feuille; x 138. j, tissu dans la partie moyenne d'une feuille; 
x 270. A;, sommet d'une feuille ; x 138. 

Fig. 15. Philonotis pergracilis. — a, plante, gr. nat. b, c, extremite de deux tiges ; x 13. d, e, f, 
feuilles; x 32. g, tissu basilaire d'une feuille; x 138. h, tissu marginal d'une feuille, vers le milieu; 
x 270. i, sommet d'une feuille ; x 138. 



Trans. Roy. Soc. Edin< Vol XLym 

Jules Cardot: Les Mousses de ^'Expedition nationals antarctique ecossaise.-Planche I. 




Fig. 6. 



M'Farlane & Ersklne, lath., Edin. 



t 



rans. Roy. Soc. Edin r Vol. XL VIII. 

Jules Cardot : Les Mousses de l'Expedition nationale antarctique e'cossaise. Planche II. 




M'Farlane & Erskine, Lith., Edin. 



«vs* MUi 



*Zm . . 



tens. Roy. Soc. Edin r y ol XLVIII. 

Titles Cardot : Les Mousses de l'Expedition nationals antarcti^ue ecossalse- Plancbe III. 




\ 

K 



\ 



Awn '* m ***$L ^Mjf 






V \ 1 \%i&'\ 

/ \ ^ \ \ ' 



^^fijl 



c. 



l/ 



Fig. 12. 




Fig. 13. 



ic^ 





Fig. 15. 








Fig. 14. 



M'Farlane & Erskine, Lith., Ediu. 



( 83 ) 



IV. — The Pharmacological Action of Harmine. By James A. Gunn, M.A., M.D., 
D.Sc. (From the Pharmacology Laboratory of the University of Edinburgh.) 

(MS received March 20, 1911. Read same date. Issued separately August 9, 1911.) 



CONTENTS. 



Introductory 

A. Lethality of Harmine .... 

B. Symptoms produced by Harmine — 

(a) in Frogs 

(b) in Rabbits 

C. Action on the Central Nervous System — 

(a) Frogs 

(b) Mammals and Pigeons . 

P. Action on Skeletal Muscle 



PAGE 

83 

84 



85 
86 



87 
87 



88 
91 
91 

94 



E. Action on the Circulation — 

(a) Heart . . . 

(6) Blood-vessels 

(c) Heart and Blood-vessels (Blood Pressure) 

F. Action on Respiration 

G. Action on Temperature 94 

H. Action on the Uterus 95 

General Summary 95 

Comparison of the Actions of Harmine and 



Harmaline 



96 



Introductory. 

The seeds of Peganum Harmala contain two alkaloids, Harmaline and Harmine. 
The pharmacological actions of the former alkaloid have been described in a previous 
communication to this Society ; # in this paper an account is given of the pharmacology 
of the second alkaloid, Harmine. 

Harmine (Ci 3 H 12 N 2 0) was discovered by Fritche in 1847. Apart from a few 
observations by Tappeiner, its pharmacology has not been investigated. Tappeiner t 
states that, in mammals at least, the general nature of poisoning by harmine is qualita- 
tively the same as by harmaline, but that the former alkaloid is weaker in action. He 
found that a dose of 0*2 gramme per kilo of harmine is fatal to the guinea-pig in about 
12 hours, while the same dose is fatal to the rabbit in about 1 hour, and that a frog is 
killed in about 7 hours by a dose of 0*03 gramme (per kilo?). He states further that 
there appears to be one qualitative difference between the actions of the two alkaloids, 
in that harmine produces paralysis of reflex excitability before arrest of the heart. It 
may be stated here that this does not constitute a qualitative difference between the 
actions of harmaline and harmine, because the same effect is produced by harmaline. 

My investigation of harmaline having shown that the actions of this alkaloid 
very intimately resemble those of quinine, a more extended investigation seemed 



* Gunn, Trans. Roy. Soc. Edin., xlvii., 1909, pp. 245-272. 
+ Tappeiner, Arehivfiir exper. Pathol, u. Pharmakologie, Bd. xxxv., 1895, p. 69. 
TRANS. ROY. SOC. EDIN., VOL. XLVII1. PART I. (NO. 4). 



13 



84 



DR JAMES A. GUNN ON 



desirable also of the pharmacology of harmine, especially as there seems some prospect 
of the alkaloids being of therapeutic value. 

I am much indebted to Dr J. F. Thorpe, F.R.S., for his great kindness in giving 
me several grammes of pure harmine for pharmacological investigation. From the 
base I prepared, according to his directions, the hydrochloride of the alkaloid, and with 
this salt all the experiments to be described were performed. 

A. Lethality of Harmine. 

The lethality of harmine was determined for frogs, guinea-pigs, rabbits, rats, and 
pigeons, with the following results : — 

Table I. — Minimum Lethal Dose by Subcutaneous Injection for Frogs. 



No. of 


Weight of 


Dose per 


Actual Dose 




Experi- 


Frog 


Kilogramme 


HI 


Result. 


ment. 


in Grammes. 


in Grammes. 


Grammes. 




1 


18 


0-4 


0-0072 


Recovery. 


2 


20 


0-5 


001 


>) 


3 


16 


06 


00096 


Death in about 6| hours. 


4 


23 


0-8 


0-0184 


>> >) )> 



Table II. — Minimum Lethal Dose by Subcutaneous Injection for Guinea-pigs. 



No. of 
Experi- 
ment. 


Weight of 

Guinea-pig 

in Grammes. 


Dose per 
Kilogramme 
in Grammes. 


Actual Dose 

in 

Grammes. 


Result. 


5 
6 

7 


620 
750 
700 


0-08 

01 

012 


05 
0-075 

0-084 


Recovery. 

>> 
Death in 2 hours. 



Table III. — Minimum Lethal Dose by Subcutaneous Injection for Rabbits. 



No. of 
Experi- 
ment. 



8 

9 

10 

11 



Weight of 

Rabbit 

in Grammes. 



1200 
1750 
1600 
1550 



Dose per 
Kilogramme 
in Grammes. 



015 
0-2 
023 
03 



Actual Dose 

in 

Grammes. 



0-18 
0-35 
0368 
0-465 



Result. 



Recovery. 

>> 
Death in 2 hours 40 minutes. 
1 hour 12 minutes. 



THE PHARMACOLOGICAL ACTION OF HARMINE. 



85 



Table IV. — Minimum Lethal Dose by Subcutaneous Injection fob Rats. 



No. of 


Weight of 


Dose per 


Actual Dose 




Experi- 


Rat 


Kilogramme 


in 


Result. 


ment. 


in Grammes. 


in Grammes. 


Grammes. 




12 


155 


01 


00155 


Recovery. 


13 


100 


015 


0-015 


)» 


14 


150 


0-2 


003 


» 


15 


100 


0-2 


0-02 


Death in 3 days. 


16 


115 


03 


0035 


„ 5-20 hours. 


17 


115 


0'4 


0-046 


„ 7-20 „ 



Table V. — Minimum Lethal Dose by Subcutaneous Injection for Pigeons. 



No. of 
Experi- 
ment. 


Weight of 

Pigeon 

in Grammes. 


Dose per 
Kilogramme 
in Grammes. 


Actual Dose 

in 
Grammes. 


Result. 


18 
19 
20 


300 
430 
420 


01 

0-12 

015 


0-03 
051 
0063 


Recovery. 

Death in 10 minutes. 



For determination of the minimum lethal dose in frogs, and for all subsequent ex- 
periments on frogs, the species Rana temporaries was used. In frogs, injections were 
made into the dorsal lymph sac, in mammals under the skin of the right flank, and in 
pigeons under the skin of the right thigh. 

From the above tables it is seen that the minimum lethal dose by subcutaneous 
injection per kilogramme is, for the frog, 0'6 gramme ; for the guinea-pig, 0*12 gramme ; 
for the rabbit, 0*23 gramme; for the rat, about 0'2 gramme; and for the pigeon, 0*1 5 
gramme. 

B. Symptoms produced by Harmine. 

(a) In Frogs. 

Experiment 3. — Rana temporaria, male, weight 16 grammes. At 12.30 p.m. the 
throat respirations were twenty in ten seconds and the cardiac impacts seven in ten 
seconds. 

At 12.40, 0*0096 gramme of harmine hydrochloride dissolved in 0*48 c.c. of saline 
solution was injected into the dorsal lymph sac. This was equivalent to 0'6 gramme 
per kilogramme. 

At 1.40 the respirations were feebler and somewhat irregular, the average rate 
being about 14 per ten seconds. The cardiac impacts were five in ten seconds and less 
distinct than before injection. The back was stiff owing to rigidity of the back 
muscles. The frog was unable to jump, and turned over with difficulty when placed on 



86 DR JAMES A. GUNN ON 

its back, an effect partly due to some stiffness of the thigh muscles caused by diffusion 
of the injected solution. The conjunctival reflex was sluggish, and there was 
some impairment of the reflex excitability of the cord as determined by electrical 
stimulation. 

At 2.30 the frog was unable to jump, and could not recover the ventral posture when 
laid on its back. 

At 3.0 the respirations had ceased and the cardiac impacts were not visible ; but, 
when the web of the foot was examined under the microscope, the blood was found to 
be circulating sluggishly. Stimulation of the skin of one leg produced no movements 
of the opposite leg, even with the coil at 30 mm. ; but stimulation of the skin over the 
sciatic nerve produced a contraction of the gastrocnemius muscle of the same side, with 
the coil at 120 mm. 

At 4.30 the circulation was found to be arrested in the web. The brain was pithed 
and the heart exposed. The heart was beating feebly at the rate of 2 in ten seconds. 
It ceased beating in ten minutes. No reflex movements could be elicited even by direct 
stimulation of the sciatic nerve with the secondary coil at 20 mm., but contraction of 
the gastrocnemius muscle of the same side was induced by stimulation of the nerve at 

100 mm. 

(b) In Mammals. 

Experiment 11. — Rabbit, 1550 grammes. At 10.55 a.m. the cardiac impacts were 
46, and the respirations 34, in ten seconds. The temperature was 99° C. 

At 11.0, 0'465 gramme dissolved in 10 c.c. of warm saline solution was injected 
under the skin of the right flank. This was equivalent to 0'3 gramme per kilogramme. 

At 11.5 there were marked tremors of the head and fore part of the body, and the 
hind limbs were extended so that the abdomen touched the ground. Three minutes 
later the tremors were more violent and the animal made frequent spasmodic movements 
forwards, the hind limbs being unable properly to support or propel the body. 

At 11.12 a slight epileptiform convulsion occurred, marked especially by clonic 
movements of the limbs, after which the animal lay quiet. A similar but more violent 
convulsion occurred a minute later, during which the animal fell on its side and rolled 
over sideways two or three times. After this the rabbit lay on its side with feeble 
pawing movements of the limbs. The respirations were 20, and the cardiac impacts 
35, in ten seconds. The temperature was 100° C. 

Up to 11.30 the symptoms were similar, but the convulsive movements became 
gradually less violent. At the end of that time the respirations were 12 in ten seconds, 
regular and deep. The cardiac impacts were palpated with difficulty, and were about 
28 in ten seconds. The skin was colder. 

At 11.50 there was almost complete motor paralysis, the animal lying continuously 
on its side and making feeble running movements occasionally. When held up by the 
ears it made no movements. Pinching the skin evoked no reflex movements, and the 
conjunctival reflex was sluggish. 



THE PHARMACOLOGICAL ACTION OF HARMINE. 87 

At 11.55 the conjunctival reflex was with difficulty elicitable. The thigh muscles 
and muscles of the right flank were stiff. The temperature was 9 6 '4° C. 

At 12.9 the animal made no movements, apart from those of respiration which were 
feeble and irregular, the rate being about 9 in ten seconds. The heart beats were feeble, 
irregular, and infrequent. 

At 12.12 the respirations ceased. The thorax was opened, and at 12.13 the heart 
found to be beating very feebly. It ceased beating in the diastolic position at 12.15. 
The muscles round the seat of injection were rigid and inexcitable. The muscles of the 
fore limbs reacted to weak stimulation of their nerves, as also did the diaphragm to 
stimulation of the phrenic nerve. 



C. Action on the Central Nervous System. 

(a) In Frogs. — The description given of the symptoms produced by lethal doses of 
harmine in the frog has shown that the chief effects referable to an action on the central 
nervous system are loss of co-ordination and of the power of jumping, arrest of respira- 
tion, and paralysis of reflex excitability. Since these effects come on at a time 
when they cannot be accounted for by paralysis of the peripheral neuro-muscular 
mechanism, they indicate that harmine paralyses the mid-brain, medulla oblongata, 
and spinal cord. 

(b) In Mammals and Pigeons. — Epileptiform convulsions form the most conspicuous 
symptom produced by large doses of harmine in warm-blooded animals. They are 
clonic in nature and are usually intermittent, intervals of quiescence ensuing between 
the convulsions. They are generally aggravated by any voluntary movement. All the 
facts observed in regard to these convulsions, among which may be mentioned their 
clonic nature, their occurrence apart from any marked increase of spinal reflex excita- 
bility, and their absence in frogs, point to their being due to an exciting action on the 
cerebrum. 

In regard to this action of harmine on the cerebrum, it is of interest to observe that 
the minimum lethal dose per kilogramme for a series of animals is roughly in inverse 
proportion to the amount of grammes of brain per kilogramme of body weight in those 
animals. A somewhat similar relationship occurs with cocaine, which also produces 
cerebral convulsions.* 

The clonic convulsions produced by harmine are not directly fatal to the animal, as 
they do not markedly interfere with respirations. When the dose is lethal, the con- 
vulsions are followed for a short time before death by a condition of motor paralysis 
due to a depressing action on the central nervous system. If the dose be not lethal, the 
convulsions are usually entirely recovered from within one or two hours. 

* Dixon, Manual of Pharmacology, 1906, p. 145. 



88 



DR JAMES A. GUNN ON 



D. Action on Skeletal Muscle. 

To ascertain the effects produced by harmine on voluntary muscle, experiments were 
made on the isolated gastrocnemius muscles of the frog, one muscle being immersed in 
a solution of the alkaloid, the other muscle in Ringer's solution. A modified Wild's 
method was employed, and to stimulate the muscles the secondary current passed 
simultaneously through both muscles. Tracings were taken on a slowly revolving drum. 

Experiment 21 (figs. 1 and 2). — Strength of solution, 1 in 2000. Normal twitches 
resulting from stimulation with break shocks are shown at 11.28, both muscles being in 
Ringer's solution. At 11.30 Ringer's solution was withdrawn from muscle B and 
replaced by a solution of harmine hydrochloride 1 in 2000 in Ringer's solution. 




Fio. 1. 



Fig. 2. 



Muscle twitches (generally three in succession) were thereafter taken at intervals ol 
two minutes, with the secondary coil at 100 mm. throughout. 

As the tracing shows, harmine causes the muscle gradually to pass into a condition 
of rigor with diminishing excitability and extent of contraction, so that in forty minutes 
the muscle had raised the lever above the level of the summit of a single twitch and no 
longer responded to the stimulus. The control muscle was unaffected. 

This effect on muscle is always produced by solutions of harmine when not less 
dilute than I in 5000, sometimes even by solutions of 1 in 10,000. Results of this 
action on muscle are exemplified in the general effects of poisoning by harmine by the 
occurrence of rigidity and impaired excitability of the muscles round the seat of injec- 
tion, and also by the unusually rapid onset of rigor mortis after lethal doses. 



E. Action on the Circulation. 
(a) Heart. 
A series of experiments was performed in which the isolated frog's ventricle was 
perfused by means of Schafer's frog-heart plethysmograph. A mixture of defibrinated 



THE PHARMACOLOGICAL ACTION OF HARMINE. 



89 



ox-blood (one part) and Ringer's solution (two parts) was used as the nutrient 
solution and as the solvent for harmine. The bulb of the plethysmograph which 
contained the heart was filled with Ringer's solution, and the contractions of the ventricle 
were recorded by means of an air-piston recorder attached by a rubber tube to the brass 
cylinder. 

Experiment 22 (figs. 3 to 5). — Strength of solution, 1 in 10,000. This strength of 



Systott 



mm 






I **' ta.ooO 



Fig. 3. 



v^nnrvnrYTTVTT 



f /2,'32, 



Fig. 4. 




Fig. 5. 



solution rapidly reduced both the rate of beat of the heart and also the amplitude of 
its excursus, the diminution of excursus being due mainly to less complete systole, 
but also to incomplete relaxation (fig. 3). Thus, in three minutes the rate fell from 18 
to 6 contractions per minute, while the excursus was reduced from 9 to 2 millimetres. 
In six minutes the ventricle was arrested in a position of almost complete diastole, and 
the normal solution was thereupon substituted for the harmine solution (fig. 4). This 
so quickly restored the heart, that in ten minutes the rate and the excursus were 
practically the same as before harmine (fig. 5). 



90 



DR JAMES A. GUNN ON 



Experiment 23 (figs. 6, 7, and 8). — Strength of solution, 1 in 25,000. The con- 
ditions and result of this experiment are detailed in the following table : — 



Table VI. — Experiment 23. 



Time. 


Kate per 
"Minute. 


Amplitude of 
Excursus. 


Solution Perfused. 


4.10 

4.20 

4.22 

4.25 

4.34 

4.44 

5.4 

5.11 

5.13 

5.16 


28 
30 

17 

13 
14 
11 

20 
26 


17 mm. 

16 „ 

1 1 mm. 
10 „ 
10 „ 
u „ 

17 ram. 
17 „ 


Normal solution. 

Harmine solution turned on (fig. 6). 

Fig. 7. 

Fig. 8. 

Normal solution turned on. 



This solution, therefore, produced considerable slowing and some weakening of the 
heart, but did not arrest it in fifty minutes. Recovery was rapid on reperfusion with 
the normal solution. 



Sy&aCz. 



I ill II f I I ( I II llllllllll 





/term***, Myfocdrljji 



Fig. 6. 




Fig. 8. 



These two experiments illustrate the chief effects of harmine on the frog's heart, 
which may now be summarised. Solutions of 1 in 10,000 or more concentrated solutions 
rapidly slow the heart and arrest it in a position of complete, or almost complete, 
diastole. Solutions of 1 in 15,000 to 1 in 30,000 produce slowing of the heart, and also 



THE PHARMACOLOGICAL ACTION OF HARMINE. 91 

some reduction in the amplitude of its excursus due to less complete systolic contraction. 
Solutions of 1 in 50,000, or weaker solutions, have no effect on the heart. 

A point of some interest, when taken into consideration with the effect of harmine 
on blood pressure in mammals, is the extreme readiness with which the heart recovers 
when the harmine is removed from the circulation. 

Other experiments have shown that the slowing of the heart produced by harmine 
is not prevented by simultaneous perfusion with atropine sulphate, and is therefore due 
to an action on the cardiac muscle. 

(b) Blood-vessels. 

To ascertain any changes produced by harmine on the blood-vessels of the frog, the 
following method was used. After the frog was pithed and the heart exposed, the 
vense cavse were cut across, and a fine cannula was tied into the left aorta, the right 
aorta being ligatured. This cannula was connected with two Marriotte's flasks containing 
the fluids to be perfused. A record was taken of the amount of fluid exuding per minute 
from the cut venae cavse. Ringer's solution was used as the normal solution and as the 
solvent for harmine. 

Perfusion of the vessels for thirty minutes with solutions of harmine of varying 
strengths gave the following results : — A solution of 1 in 1000 reduced the flow from 
1*8 c.c. per minute to 0"9 c.c. per minute; a solution of 1 in 2500 reduced the flow 
from 24 c.c. per minute to 17 c.c. per minute; a solution of 1 in 5000 reduced the 
flow from 27 c.c, per minute to 2*0 c.c. per minute ; while a solution of 1 in 7500 had 
no effect on the flow through the vessels. Harmine has therefore a slight constricting 
action on the frog's blood-vessels. 

(c) Heart and Blood-vessels. (Blood Pressure.) 

In all blood-pressure experiments the animals (rabbits or cats) were first anaesthet- 
ised with chloroform ; the trachea was then exposed, and a cannula tied into it through 
which diluted ether was thereafter inhaled. A cannula in the left carotid artery was 
connected with the manometer. Respirations were recorded by means of a double 
stethograph attached by a band round the thorax and connected with a Marey's 
tambour. Injections were made into the right jugular vein. It was found in pre- 
liminary experiments that the minimum lethal dose of harmine injected in this way is 
about 0"03 gramme per kilo. 

Experiment 24 (Table VII., figs. 9 and 10). — Rabbit, 2400 grammes. The first 
injection of 0*01 grm. per kilo, equivalent to one-third of the intravenous minimum 
lethal dose, produced a somewhat transient fall of blood pressure, the normal level 
being restored in about ten minutes. The fall of pressure was accompanied by a 
slowing of the heart, which was evidently, in part at least, a causal factor. Recovery 
of blood pressure occurred in spite of further slowing of the heart. It is noteworthy 

TRANS. ROY. SOC. EDIN., VOL. XLVIIL, PART I. (NO. 4). 14 



92 



DR JAMES A. GUNN ON 





THE PHARMACOLOGICAL ACTION OF HARMINE. 



93 



that, though the dose injected in this case was so large as one- third of the minimum 
lethal, no effect was produced on respiration. 

The second injection produced a rapid decline of blood pressure, accompanied by- 
slowing, and later by great feebleness, of the heart's contractions. Respiration was not 
much affected until the blood pressure reached a very low level. This, together with 
the facts that respirations continue as long as the heart beats and that respirations are 
unaffected except by doses which produce a grave fall of blood pressure, goes to show 
that death from intravenous injection of harmine is mainly, if not solely, due to cardiac 

failure. 

Table VII. — Experiment 24. 



Time. 


Dose of 
Harmine 


Average B.P. 


Pulse 
Rate 


Rate of 
Respirations 


Respiration 


Notes. 




intravenously. 


in mm. 


per 10 
seconds. 


per 10 
seconds. 


Excursus. 




11.37 




100 


48 


10 


2 mm. 




11.38 


0"01 grm. per kilo. 




• • • 






Fig. 9. 


11.38.30" 




82 


42 


10 


2 mm. 




11.40 




92 


46 


10 


2 




11.48 




98 


40 


10 


2 „ 




11.50 


0*03 grm. per kilo. 










Fig. 10. 


11.50.30" 


. . . 


60 


37 


7 


2 mm. 




11.51 


. . • 


50 


14 


5 


1 „ 




11.52 




35 


16 


3 


o-5 ;, 


Pulse waves very 
small. 


11.54 


. . • 


28 


10 


3 




)> 


11.56 


... 
















The effects of harmine on blood pressure, as deduced from this and other experiments, 
may be briefly summarised. Apart from very small doses, which sometimes produce 
an insignificant rise of blood pressure, the chief action of harmine is to produce a fall of 
blood pressure, due to slowing of the heart, and, in the case of lethal doses, also to 
enfeeblement of the heart's contractions. The slowing of the heart is not prevented by 
previous administration of a dose of atropine sufficient to paralyse the vagal endings, 
so that it is due to an action on the cardiac muscle, as was found also in the case of the 
frog's heart. 

The only point of importance which remains to be discussed is whether the fall of 
pressure is due solely to cardiac causes or is due partly to vascular dilatation. To 
determine this, several experiments were made in which a record was taken of the 
blood pressure and also of the volume of the kidney or of a loop of intestine. 

Experiment 25 (fig. 11). — Cat, 2700 grammes. Blood pressure was recorded as 
in the previous experiment, and the kidney volume was recorded by an oncometer and 
air-piston recorder. An injection of 0'004 gramme per kilo was given, about one- 
eighth of the minimum lethal dose. This reduced the blood pressure in one minute 
from 120 mm. to 88 mm., and the pulse rate from 18 to 13 per ten seconds. There 



94 



DR JAMES A. GUNN ON 



was meantime a considerable reduction in the volume of the kidney, showing that there 
was no dilatation of its blood-vessels. Similar results were obtained with the 




Fig. 1!. 

intestinal volume. It would appear, therefore, that the fall of blood pressure is due 
chiefly to diminished output from the heart, and not to dilatation of the abdominal 
vessels, though experiments on the intact animal gave indications of some dilatation 
of the vessels of the skin. 

F. Action on the Respiration. 

Lethal doses of harmine paralyse the respiration in frogs a short time before 
cessation of the heart beats, but at a time when there is great feebleness of the circula- 
tion. In mammals the same effect is obtained with small lethal doses ; but with 
rapidly lethal doses, especially if intravenously injected, the heart beats and respirations 
cease at the same time. At the time of death the diaphragm reacts to weak stimulation 
of the phrenic nerve, and respiratory failure is probably partly due to a direct depressing 
action of harmine on the respiratory centre, and is partly consequent upon circulatory 
failure. 

In the unansesthetised mammal sublethal doses produce a distinct stimulation of 
respiration. This does not occur if the animal is anaesthetised with chloroform or 
ether, a very common phenomenon with respiratory stimulants. 



G. Action on Temperature. 

Large doses of harmine cause a fall of temperature in mammals, an effect which 
has been shown by Harnack * to be generally true of convulsant poisons. The fall of 
temperature is sometimes preceded by a slight transient rise. 

* Harnack, Archiv fur exper. Path. u. Pharmakol., 1897, Bd. xxxviii. 



THE PHARMACOLOGICAL ACTION OF HARMINE. 



95 



H. Action on the Uterus. 

Rabbits were used for these experiments. They were anaesthetised as for blood- 
pressure experiments and kept during the experiment in a bath of saline solution at 
38° C, enough of the body being submerged to ensure that the uterus was never 
exposed to the air. The abdomen was then opened in the middle line, and the uterus, 
isolated from the surrounding viscera, was connected with a lever writing on a slowly 
revolving drum. 

Experiment 26 (figs. 12 to 14). — Rabbit, 2350 grammes, parous, non-pregnant. 
Slight spontaneous contractions occurred regularly at the rate of about 2 per minute. 




H*m4»l HA. 
I/O •'*« /«•>• /4*r Hi'* 



Fig. 12. 



At 11.0, 0*002 gramme per kilogramme of harmine hydrochloride was injected intra- 
venously, and this produced a marked augmentation of the uterine contractions, which, 
however, soon passed off. At 11.15 a second injection of twice the former amount was 





Fig. 13. 



Fig. 14. 



given, and this produced very powerful uterine contractions (fig. 13), which were 
repeated at intervals afterwards (fig. 14). The resting tone of the uterine muscle was 
also increased. 

It is evident, therefore, that harmine exerts a decided stimulating effect on the 
uterine contractions. 

General Summary. 

The minimum lethal dose of harmine hydrochloride per kilogramme by subcutaneous 
injection is, for the frog, 6 gramme ; for the guinea-pig, 0'12 gramme ; for the rabbit, 
0'23 gramme ; for the rat, 0'2 gramme ; and for the pigeon, 0*15 gramme. 

In frogs, lethal doses of harmine paralyse the mid-brain, medulla oblongata, and 
TRANS. ROY. SOC. EDIN., VOL. XLVHL, PART I. (NO. 4). 15 



96 THE PHARMACOLOGICAL ACTION OF HARMINE. 

spinal cord. Abolition of reflex excitability occurs before arrest of the heart, and 
before paralysis of the voluntary muscles. In mammals large doses cause epileptiform 
convulsions, cerebral in origin. If the dose be non-lethal these are soon recovered 
from ; if the dose be lethal, they give place to a condition of paralysis of the central 
nervous system, which endures for a short time before death. 

Harmine produces rigor and inexcitability of an isolated muscle, but, even with 
lethal doses, the concentration of harmine in the blood is not sufficient to render this 
action of importance in the general effects of this alkaloid. 

Strong solutions of harmine perfused through the frog's heart slow the heart and 
arrest it in a position of almost complete diastole ; weaker solutions slow the heart and 
diminish the completeness of systolic contraction. The effects are due to an action 
on the cardiac muscle. 

Harmine has a slight peripheral constricting action on the frog's blood-vessels. 

In mammals, harmine, in doses which have any marked effect on blood pressure, 
produces a fall of blood pressure due chiefly to slowing, or, in the case of large doses, 
to slowing and weakening, of the heart's contractions. Cardiac failure is the chief 
cause of death from harmine poisoning. 

Sublethal doses of harmine stimulate respiration ; lethal doses paralyse respiration, 
partly from a direct action on the respiratory centre and partly as a consequence of 
circulatory failure. 

Like many convulsant poisons, harmine in large doses produces a fall of temperature 
in mammals. Even in small doses it stimulates the contractions and augments the 
tone of uterine muscle. 

Comparison of the Actions of Harmine and Harmaline. 

The pharmacological actions of harmine resemble very closely those of harmaline in 
so far as the symptoms produced in the intact animal and the effects produced on the 
various systems and on isolated tissues are qualitatively the same in the case of both 
alkaloids. For this reason the pharmacology of the former alkaloid has been discussed 
more briefly. Harmine is, however, only about half as toxic as harmaline ; and probably 
the chief reason for the relatively lower toxicity of harmine is that the primary stimulat- 
ing action on the central nervous system less readily gives place to paralysis, and 
hence respiratory paralysis plays a less important part in the production of its lethal 
effects than is the case with harmaline. 

As the alkaloids can easily be obtained from the seeds in a mixed form, whereas 
their separation from one another is, I understand, a difficult and tedious process, this 
close similarity in their pharmacological actions possesses this importance, that the 
mixed alkaloids would apparently be as effective therapeutically as either alkaloid 
alone, should a therapeutic use for them be found. 



( 97 ) 



V. — On the Resistance to Flow of Water through Pipes or Passages having 
Divergent Boundaries. By Professor A. H. Gibson, D.Sc., University College, 
Dundee. Communicated by Professor W. Peddie, D.Sc. 

(MS. received March 20, 1911. Read July 3, 1911. Issued separately August 30, 1911.) 

CONTENTS. 



PAGE 

1. Introduction 97 

2. Circular Pipes with Uniformly Diverging 

Boundaries 98 

3. Effect of Projecting Smaller Pipe into Space 

bounded by Diverging Walls . . . 102 



PAGE 

■4. Rectangular Pipes with Uniformly Diverging 

Boundaries 103 

5. Pipes of Square Section with Uniformly 

Diverging Boundaries .... 105 

6. Form of Pipe for Minimum Loss of Head . 106 

7. Summary and Conclusions . . . .113 



§ 1. Introduction. 

Some time ago the author published # the results of a series of experiments on the 
flow of water through tubes having uniformly diverging boundaries, in which the loss 
of energy corresponding to given angles of conicity of the tubes was determined. These 
pipes, some circular, others square or rectangular in cross section, had the same initial 
and the same final areas, these being respectively the same as those of circles of 1'5 
inch and 3"0 inches diameter, the ratio of initial to final mean velocity of flow being 
in each case 4 to 1. 

The present experiments were planned with a view of extending the investigation 
to cover a series of- values of the ratio of enlargement, and also of ascertaining whether 
the loss in similar pipes having identical mean velocities of flow varies with or is in- 
dependent of their dimensions. 

In some cases where such divergent passages form an essential feature of a hydraulic 
machine it is the practice to project the boundaries of the smaller pipe for some distance 
into the space bounded by the divergent walls, and the work has been extended to 
investigate the effect of this method of construction on the efficiency of energy trans- 
formation. 

The question as to the shape of pipe giving the least loss of energy with given 
initial and final areas and with a given length — a problem of much practical import- 
ance — was also touched upon in the former paper, and further investigation has been 
directed to this point. 

Altogether upwards of ninety pipes have been examined. The experiments have 
been carried out in the engineering laboratories of University College, Dundee, the 
apparatus used and the methods of making measurements and of carrying out the 
work being substantially illustrated and described in the former paper.t Of the pipes 
examined those of rectangular section were of wood carefully made to template and 

* Proc. Roy. Soc, A, vol. Ixxxiii., 1910, p. 366. t Ibid., p. 368. 

TRANS. ROY. SOC. EDIN, VOL. XLVIII. PART I. (NO. 5), 16 



98 



PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 



varnished. The circular pipes were of brass bored out to the correct taper, or, in the 
case of the pipes with varying curvature, carefully bored out to template. Calling 6 
the angle of conicity of a pipe (the angle contained between its opposite faces), 
the following table shows the range of uniformly tapering circular pipes experi- 
mented upon : — 



Initial Diameter. 


Final Diameter. 


Ratio of Final to 
Initial Area. 


Value of 6. 


Inches. 
•65 


Inches. 
215 


1096 


180° (sudden enlargement) 
10°, 20°, 40°, 60°, 90°, 180° 


•50 


1-50 


9-0 


100 


3-00 


9-0 


20°, 40°, 60°, 90°, 180° 


*l-50 


3-00 


4-0 


3°, 4°, 5°, 7!°, 10°, 12f , 15°, 171°, 
20°, 30°, 40°, 50°, 60°, 90°, 180° 


2-00 


3-00 


2-25 


10°, 20°, 40°, 60°, 90°, 180° 



The mean velocities of flow in these experiments ranged from 1'83 feet per second 
to slightly over 21 feet per second. The results show that in any given pipe the loss 

I si) 4) \- 

of head, expressed as a percentage of the loss, ^J— — 2z_ , theoretically obtained at a 

sudden enlargement between the same areas, does not vary in any definite manner with 
the velocity, and is, in fact, sensibly constant for all velocities above the critical ; and 
in giving the results the losses have, in every case, been expressed as a percentage 
of this quantity. 

§ 2. Experiments on Circular Pipes with Uniformly Diverging Boundaries. 

The following table gives the mean of all experiments carried out on each pipe, 
while the results are shown graphically and are compared with those of the former 
experiments, in fig. 1 : — 



(v — ■ V ) 
Loss of Energy expressed as a Percentage of ^-1— — ^- , the Theoretical Loss at a 

Sudden Change of Velocity from v 1 to v 2 . 



Value of 6 



Pipe diameters 

in inclies. 



•65 to 2-15 

•50 „ 1-50 

1-0 „ 3-0 

1-5 „ 30 

2-0 „ 3-0 



180°. 



103-5 

102-8 

102-1 

10P7 

99-2 



90°. 



102-6 
104-1 
111-1 
112-1 



60*. 



102-8 
101-3 
1205 



40°. 



82-0 

80-8 

101-7 

88-7 



20°. 



45-0 
44-0 
42-5 
41-9 



10° 



16-6 

17*5 

18-6 



* These experiments are described in the former paper. 



THROUGH PIPES OR PASSAGES HAVING DIVERGENT BOUNDARIES. 



99 




Bz 



Vl) 



JO 82B}U80J8d V SB p3SS9JClx8 pB9lj JO SSO - ] 



100 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 



From these results it appears that the total loss in the pipes having 9= 180° is in 
every case in close agreement with that given by Boulanger's formula, ^— 1 ^-. The 

actual loss is, in the majority of cases, slightly greater than that given by this formula. 
Experiments on larger pipes having a sudden change in diameter from 3 inches to 6 
inches and from 4 inches to 6 inches respectively * show an experimental loss some- 
what less than that given by the formula. The percentage loss increases slightly with 
the area-ratio, and in pipes with the same area-ratio is greater the smaller the pipe. 

Denoting the ratio of enlargement by m, and the smaller diameter by d, the loss 
at a sudden enlargement for values of m between 2 and 12, and for pipe diameters 
ranging from '50 inch to 6 inches, can be expressed, within narrow limits, by the 
relationship — 

loss of head = l_02^25m - 2W | (,, - t,,)» ( ^ 



100 



2g \ 



The following table shows the results obtained by the use of the formula, against 
those experimentally obtained : — 



Size of Pipes. 


Value of m. 


(rtf y \2 

Loss expressed as Percentage of V -J— — ^- . 

2<7 


Experimental. 


By Formula. 


•65 to 2-15 inches 
•50 „ 1-5 „ 

10 „ 3-0 „ 

1-5 „ 3-0 „ 

20 „ 3-0 

3-0 „ 6-0 

40 „ 6-0 


10-96 
9-0 
90 
4-0 
2-25 
4-0 
2-25 


103-5 
102-8 
102-1 
101-7 

99-2 

97-5 

92 (approx.) 


103-9 

103-7 

102-8 

100-5 

99-1 

97-5 

95 



As 9 is diminished from 180° the percentage loss in every case increases, attains 
a maximum value for some value of 9 in the neighbourhood of 65°, and afterwards 
diminishes rapidly with 9 until 9 is about 5° 30'. This value gives a minimum loss of 
approximately 13 "5 per cent. Any further diminution in 9 is accompanied by an 
increased loss owing to the large value of the wall friction in pipes of the comparatively 
great length accompanying such small values of 9. The value of 9 which coincides 
with the maximum loss, varies somewhat, both with the size of pipe and with the 
area-ratio, increasing slightly with the dimensions when the latter is constant, and 
with the area-ratio when the mean pipe diameter is constant. Over the range of 
pipe diameters and area-ratios examined in these experiments its value lies between 
63° and 70°. 

* Brightmore, Proc. Inst. Civil Engineers, vol. clxix., 1906-7, Pt. iii. p. 322. Here the loss of head was 97 5 per 
cent, of the theoretical for an enlargement of area 1 to 4, and was about 92 per cent, of the theoretical for an 
enlargement of 1 to 2-25. 



THROUGH PIPES OR PASSAGES HAVING DIVERGENT BOUNDARIES. 



101 



The maximum percentage loss varies in these experiments between 105 per cent, 
and 122 per cent. It increases with the mean diameter of the pipe for the same area- 
ratio, and also increases as the ratio of enlargement is reduced. 

As 6 is diminished below 60° the percentage loss curves rapidly converge, and for 
values of 6 less than 25° these are very sensibly the same for all the pipes. With 6 = 20° 
the percentage losses due essentially to the divergence of the walls, the friction losses 
being calculated as explained in the former paper.* are as follows : — 



Pipe diameters (inches) . 


■50 to 1-50 


1-0 to 30 


1-5 to 3 


2-0 to 3-0 


Percentage loss 


410 


41-7 


40-5 


40-3 



For a given ratio of enlargement, for values of less than 60°, the percentage loss 
increases slightly as the mean diameter diminishes, and, for a given mean diameter, on 
the whole increases slightly although irregularly as the ratio enlargement is reduced. 
The value of 6 giving rise to the same loss as is experienced with a sudden change of 
section, varies within fairly wide limits from 41° to 60°, being slightly greater, for a 
given area-ratio, the larger the pipes, and, for a given mean diameter, increasing in an 
irregular manner with the area-ratio. Its mean value over the range of ratios con- 
sidered is approximately 50°, and where, in the design of hydraulic machinery, it is 
necessary for this value to be exceeded, a sudden enlargement of section will give a more 
efficient transformation of energy than will a uniformly tapering pipe. For values of 
between 7 '5° and 35° the loss may be expressed with a fair degree of accuracy by the 
relationships — 

loss = "0110 1 ' 22 i-J y~ feet, where is in degrees, 



or 



2<? 



!,,:.: :-; oo ( tan- Y* ^ ^ feet. 
2/ %i 



This latter relationship becomes of importance in the design of trumpet-shaped 
pipes to give a minimum loss of energy. Values obtained by calculation from these 
formulas are compared in the following table with those obtained experimentally : — 



Percentage Loss. 


6°. 


Mean of 

Experimental 

Values. 


By Calculation. 


1-10 122 . 


350(tanf Y". 

12-8 
181 

29-9 
42-4 
562 
73-0 
85-7 


7-5 
10 
15 
20 
25 
30 
35 


145 
175 

28-0 
435 
58-0 
71-0 
80-8 


12-8 
18-2 
29-7 
42-5 
55-5 
70-5 
83-8 



* Ibid., p. 370. 



102 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 



§ 3. Effect of projecting Smaller Pipe into Space bounded 
by Divergent Walls. 

For the purpose of examining this effect, a thin sleeve of internal diameter '90 
inch and external diameter l'O inch was prepared and used inside the various I -inch 
pipes, as shown in fig. 2. 

The length / of the internal projection was varied by varying the thickness of the 
washer W. The pressure in the smaller pipe was measured at two points distant 
respectively 12 inches and 20 inches from the open end, and the pressure at this end 
was deduced from these readings on the assumption that the friction loss per unit 




Fig. 2. 



length of the pipe is constant. The percentage losses in a conical pipe with initial 
and final diameters "90 inch and 3"0 inches respectively are very sensibly the same as 
in one with initial and final diameters TO inch and 3'0 inches, and the losses actually 
observed with the projecting pipes are therefore given, for comparison with those 
given by the latter pipes. 

The results of the experiments are as follows : — 



Percentage Loss of Head. 


Value of . 


40° 


60° 


90° 


Pipe with diameters 
1 in. and 3 in. 


80-8% 


10L3% 


104-1% 


Projecting pipe with 
diameters - 90 in. 
and 3 in. 


Pipe projecting — 

•50 in. . . 98-0% 
LOO in. . . 108-0% 


Pipe projecting — 
•25 in. . . 104-5% 
•50 in. . . 107-2% 

L00 in. . . 112-0% 


Pipe projecting — 

•25 in. . . 107-7% 



THROUGH PIPES OR PASSAGES HAVING DIVERGENT BOUNDARIES. 



103 



In every case the effect of the projection is to increase the loss, the effect being 
greater the greater the length of the projection and also the less the angle of divergence 
of the conical sides. The loss is, in fact, greater in every case than would be experienced 
if the pipe walls were tapered off from the extremity of the projecting pipe, as shown 
by the dotted lines in fig. 2. 



§ 4. Rectangular Pipes with Uniformly Diverging Boundaries. 

Three sets of rectangular pipes were examined, these having one pair of sides 
parallel and 1*329 inch apart in every case. The areas of the small and of the large 
ends of these pipes were, in the case of the pipes having area-ratios of 4 : 1 and 9:1, 
identical with those of the circular taper pipes having the same ratios of enlargement. 
The details of the pipes are as follows : — 



Dimensions of Pipes (inches). 


Area-Ratio. 


Values of 6. 


Small End. 


Large End. 


•590 x 1 -329 
1-329x1-329 
1-329x1-329 


5-315 x 1-329 
5-315x1-329 
2-990x1-329 


9 : 1 

4 : 1 

2-25 : 1 


10°, 15°, 20°, 26°, 40°, 60°, 90° 
5°, 10°, 15°, 22° 42', 30°, 40° 
10°, 15°, 20° 30° 



The results of the means of the experiments on these pipes are plotted in fig. 3. 
From these it appears that the percentage loss in such pipes is very approximately the 
same for all ratios of enlargement between 2 '2 5 to 1 and 9 to 1 for values of 6 between 
10° and 40°, and that it varies but little with the dimensions of the pipe. The minimum 
loss is obtained when 6 is approximately 11°, the percentage loss under these circum- 
stances being about 17 "5 per cent. As is increased the loss increases rapidly, and 
attains a value of 100 per cent, when 6 is between 31° and 40°, the value of this critical 
angle being less with the smaller ratios of enlargement and with the pipes ha vino- the 
smaller mean sectional areas. For values of 6 between 10° and 35°, the only values of 
any use in practice, the loss can be expressed with a fair degree of accuracy by the 
relationships — 

loss = 0072 6 vm ("l"^)' feet, where 6 is in degrees. 



■ 5-30 U^T^lZ^ feet. 
\ 2/ 2(/ 



The following table shows a comparison between experimental results and values 
calculated from these relationships : — 



104 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 



60 

c 

CD 



o 



14.V 




























HO 






















































»oo 
























''A 


'y> 




























90 






















K- Ay? 
































80 




















#/ 


7 
























T/ / / 

//// 

// / / 








70 




















If' 

I / 

r / / 




















>»/ 




^/// 


/ / 








Go 














••/ 




••/// 
# 


1 




















kit 
7 




u lll / 










5o 












9/ 

to/ 






1 1 ^- 




































y\ e> 
















// / 


V 










*J-° 
















f- S 6 


































































w 




















M ' 










§ / 












So 








# / 
































/'/ 1 


























r* 














// 


'^y " *¥ • * 


































// 


) 4 


























2o 


\ 












• • Soup RE.. 4-: 1 
























































\ \ 












' /" 












S 














































/o 




































































/o' 



20° 



3o« 



^ 



s 



3E 



>/0 -2o .30 .4.0 .50 

Angles between divergent sides of pipes. 



'60 



4-O e 



•yo RfljsiRNs 



Fig. 3. — Percentage Loss of Head in Straight Taper Fipcs of Square and Rectangular Section. 



THROUGH PIPES OR PASSAGES HAVING DIVERGENT BOUNDARIES. 



105 





Percentage Loss of Head. 




8. 
10° 


Experimental 
(mean). 


•720 1 ' 4 . 


530 ( tan :J . 


17-6 


18-0 


17-8 


15° 


29-5 


317 


313 


20° 


47-5 


47-5 


46-7 


25° 


69-5 


64-8 


64-2 


30" 


90-0 


850 


83-5 


35° 


103-0 


104-0 


103-5 



A comparison with circular taper pipes of the same initial and final areas and having 
the same values of 6 shows, as indicated in the following tabic, that the rectangular 
pipe gives a greater loss of head, except for values of between 10° and 15° : — 



Percentage Loss of Head. 


9. 


Circular Pipe (mean). 


Rectangular Pipe (mean). 


10° 
20° 
30° 
40° 


17-5 
43-5 
71-0 
89 


176 

47-5 

90-0 

108-0 



If pipes of the same length and same ratio of enlargement are compared, however, 
the rectangular pipe, having a greater value of 9 than the circular pipe, is much 
less efficient except for such lengths as would make #, in the rectangular pipe, 
approximately 11°. 

Experiments on rectangular pipes having an enlargement ratio of 9 to 1, and having 
6 respectively 40°, 60°, and 90°, showed corresponding losses of 115, 122*3, and 119 
per cent. This indicates a maximum loss when 6 is approximately 70°, or when it has 
sensibly the same value as gives maximum loss in the corresponding circular pipe. 



§ 5. Pipes of Square Section with Uniformly Divergent Boundaries. 

Two pipes, having a smaller section T329 inch square and a larger section 2'658 
inches square and with 9 respectively 7 "5° and 30° were examined, these completing 
the series examined in the preceding experiments. The results of the whole set of 
experiments are plotted in fig. 3. The percentage loss is a minimum for a value of 9 
in the neighbourhood of 6", practically the same as for a circular pipe, and has a value 
of about 14 "5 per cent. 

For larger values of 9 the loss in a square pipe is considerably greater — 85 per cent. 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 5). 17 



106 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 

greater when = 20' — than in the corresponding circular pipe, while for values of 6 
greater than 8° it is also greater than in a rectangular pipe with the same value of 6 
but having one pair of sides parallel. 

§ 6. Pipes of Best Form. 

Where no restrictions are placed on the length of the pipe, a straight taper pipe 
having a divergent angle of about 6° in the case of a circular pipe and of a square pipe, 
and of about 11° in the case of a rectangular pipe, will give the minimum loss of energy 
between inlet and outlet.* The great length of such a pipe, however, renders its use 
impossible in many cases which occur in practice, and in such a case it becomes im- 
portant to determine what form of passage will give the least loss for a given length 
and given ratio of enlargement. It would appear that such a pipe should be trumpet- 
shaped, the angle of divergence being least at the small end of the pipe where the 
velocity is greatest, and gradually increasing as the velocity diminishes. 

While impossible to determine the best curve from purely a priori reasoning, it 
seemed reasonable to suppose that the loss might be least either 

(1) with a pipe giving uniform retardation (— = constant) , 

(2) ., ,, ,, change of velocity per unit length of pipe ( — = constant ) , 

\dx J 

(3) ,, ,, ,, loss of head per unit length of pipe. 



or 



or 



In the former experiments pipes were prepared both of circular and of rectangular 
section for the purpose of testing the validity of the first two of these assumptions.! 
In the case of the rectangular pipes, which had the same initial and final areas and the 
same length as a straight taper pipe having 0=22° 42',} it appeared that the loss was 

reduced by 5'3 per cent, in the former case (-? = constant) and by 12'1 per cent, in 

-j- = constant). In the case of the circular pipe, however, the loss was 

actually greater in the pipe having -=- constant than in the straight taper pipe. On 

this account it was decided to test the validity of the third of these assumptions, and 
the pipes referred to in the table opposite were prepared for comparison with straight 
taper pipes of the same length and same change of section. 

The boundary curves for these pipes were set out from equations deduced as follows: — 
The loss in a straight taper pipe whose angle of divergence is is proportional to S(vf 

tan -J where n= 1-40 for a rectangular pipe. Hence in 

* This statement requires modification in view of considerations outlined at a later stage of the paper. 

t Ibid., pp. 367, 375, 376. 

I Not 20° as stated in the former paper. 



THROUGH PIPES OR PASSAGES HAVING DIVERGENT BOUNDARIES. 



107 





Areas. 


Ratio of 
Areas. 


Length. 


Angle of 
Corresponding 

Straight 
Taper Pipes. 


Initial. 


Final. 


Rectangular 
Pipes. 

V 


•590 in. x 1-329 in. 


5-315 in. x 1329 in. 

>> ii 


9 
9 


1 

1 


ins. 
1340 
10-25 


20° 
26° 


1-329 in. x 1-329 in. 

>' )! 
>) 11 


5-315 in. x 1-329 in. 

11 J! 
)) )) 


4 

4 
4 


1 

\ 


9-94 
7-45 
5-49 


22° 42' 

30° 

40° 


1-329 in. x 1-329 in. 

ii ii 


2-99 in. x 1-329 in. 

II 19 


2-25 
2-25 


1 
1 


4-70 
310 


20° 
30° 


Circular « 
Pipes. 


"50 in. diam. 


1*50 in. diam. 
>> 


9 
9 


1 

1 


3-80 
1-870 


15° 
30° 


1"50 in. diam. 


3 - in. diam. 
>> 


4 
4 


1 
1 


8-575 
2-802 


10° 
30° 


20 in. diam. 


3'0 in. diam. 


2-25 
2-25 


1 
1 


5-715 
1-870 


10* 
30° 



a length 8x of a trumpet- shaped rectangular pipe, over which the mean angle is 0, the 

loss is presumably proportional to S(v) 2 (-f^) or to -\=-^ . (-p) • &» where y is the 

half-breadth of the pipe and where x measures the distance of the element under con- 
sideration from some datum point on the axis of the pipe. 
But in such a pipe v 2 oc?/~ 2 , 

.-. loss in length 8a; = _ (y~ 2 )( — ) 8x . 
dx \dxJ 

For this to be constant per unit length 

^ (2/ ""XI) 14=constant ' 



or 

y-tf _i ) = constant , 
\dxj 

dx 

••• IW>' 

J>J\ 

.: yi -*-y-~=K(*-*d (1) 

If the origin from which x is measured be taken at the small end of the pipe where 
the half-breadth is y^, x x = 0, and if I be the length of the pipe and if y 2 be the half- 
breadth at the larger end, x 2 — x 1 = I, and 



108 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 



from which K can be calculated. Knowing K, the value of y corresponding to any 
value of x can be readily obtained from equation (1), and the rectangular pipes were 
made to templates representing curves calculated in this way. 
In a circular pipe n = T22, while v 2 °cy~\ so that 



L («-«)(^Y"= constant. 



^^ ^ 




^'A.v.^^k*.^' 



Fig. 4. 



(2) 



Proceeding as above, this gives 

Z/r 1 ' 25 - Z/ -1 ' 25 = K(.x- - a;i ) .... 
and on the same assumptions as before 

K -i{yr 1 *-«T w, f 

l being the length, and y 1 and y 2 the smaller and larger radii respectively. The circular 
pipes were made to templates set out from this equation. 

Where the length of pipe is great, or the ratio of areas small, the curves thus formed 
may, at the smaller end of the pipe, diverge at an angle less than that (6° in a circular- 
pipe and 11° in a rectangular pipe) giving minimum loss, and in such a case the pipe 



THROUGH PIPES OR PASSAGES HAVING DIVERGENT BOUNDARIES. 



109 



would be made to diverge uniformly at this best angle up to a point where its straight 
sides would intersect the calculated curved sides. 

If, on the other hand, the length of pipe is small or the ratio of areas large, the 
calculated curves may, towards the larger end of the pipe, diverge at an angle greater than 
that (about 35° in a rectangular pipe) giving a loss equal to that at a sudden enlargement. 

In such a case — as was confirmed by experiment — a more efficient pipe is obtained 
by enlarging the pipe to its final section by a sudden enlargement at the point at which 
the angle of divergence becomes equal to this critical value. 

A still more efficient pipe is obtained if, from the point at which the angle of 
divergence becomes 35°, the section is enlarged gradually, the best angle of divergence 
being found to vary but slightly in such circumstances and in any cases likely to be 
found in practice, being approximately 20°. 

All the pipes were first constructed with a continuous curve from inlet to outlet, 
and those pipes in which the angle at outlet exceeded the critical value were, after 
being used, modified as shown by the dotted lines in fig. 4, A. After further examination 
they were again modified as shown in fig. 4, B, with the results indicated in the following 
table : — 





Percentage Loss in Trumpet-shaped Pipes. 






( 

Rectangular 
Pipes. 

• 

Circular 
Pipes. 

1 


Ratio of 
Enlargement. 


Value of 6 

in Straight 

Pipe of 

same Length 

as Curved 

Pipe. 


Straight 
Pipe. 


Curved Pipe. 


Continuous 
Curve. 


Modified 
as in "A." 


Modified 
as in " B." 


9 : 1 
9 : 1 


20° 
26° 


48-0 
795 


24-1 
37-6 


21-3 
34-2 


19-5 
30-3 


4 : 1 

4 : 1 
4 : 1 


22° 42' 

30° 

40° 


505 

85T 
100-0 


(47-6* 

{ 44-4 f 

(26-4 

38-0 

48-9 


35-9 

42-7 


339 
41-5 


2-25 : 1 
225 : 1 

9 : 1 
9 : 1 


20° 
30° 


47-4 
94-0 


28-1 
53-0 






15° 
30° 


28-0 
67-0 


23-4 
394 


38-0 




4 : 1 

4 : 1 


10° 
30° 


17-5 

825 


13-9 

38-8 


35-5 




2-25 : 1 
225 : 1 


10° 
30° 


18-5 
69-0 


14-5 
40-9 







* — = constant. 
dv 



t — = constant. 
dv 



110 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 

From these results it appears that by the use of curved pipes of this form the loss 
with a given length of pipe may be reduced considerably as compared with the loss in 
a straight taper pipe between the same limits of area. The proportional saving is 
greater the greater the ratio of enlargement and also the shorter the pipe. 

While no direct evidence that this is the best possible curve has been obtained, the 
great reduction in loss accompanying its use is evidence that no great improvement by 
a modification in its equation is probable, or indeed possible. 

For values of less than about 15° in the case of a rectangular pipe and less than 
about 7 '5° in the case of a circular pipe, evidently little is to be gained by introducing 
the curve. The proportional saving following on its introduction varies in these experi- 
ments from 62 per cent, in the case of a rectangular pipe having an enlargement ratio 
of 9 : 1 and corresponding to a straight taper pipe with 6 = 26°, to 22 per cent, in the 
case of a circular pipe with a ratio of 2"25 : 1 and corresponding to a straight pipe 
with = 10°. As might be expected, the gain is more marked in a rectangular pipe, 
in which the enlargement of section takes place in one plane, than in a circular pipe. 
The effect of modifying the outlet end of the pipe as indicated in fig. 4 (A and B) is 
somewhat surprisingly large, the mere cutting away of the curved boundary to form 
a sudden enlargement reducing the percentage loss by about 7 per cent, on the 
average. The pipe as thus modified is, in effect, moreover, shorter than the original 
curved pipe giving the same change of section, and this led to the examination of a 
further series of pipes designed from considerations based on the following reasoning. 

The loss of head in a pipe whose section increases gradually from A x to A 3 , and 
which then suffers a sudden enlargement of area to A 2 , might be expected, on 
theoretical grounds, to be equal to the sum of the separate losses which would be 
experienced in the taper portion of the pipe and at the sudden enlargement, if these 
were independent of each other. By reducing the angle of divergence of the first 
portion of such a tube, the sudden enlargement of section and the accompanying loss 
is made greater, but the loss in the diverging portion is reduced in a double degree, 
since not only is the numerical coefficient expressing such loss as a percentage of 
(^i — v zfl^9 diminished, but A 3 is diminished at the same time, and thus the factor 
v l — v 3 is also diminished. A diminution in the angle of divergence therefore causes 
a rapid diminution in this portion of the loss, which may, or may not, be counter- 
balanced by the loss at the sudden enlargement of section. Owing to the comparatively 
low velocities at the large end of the pipe, however, except in pipes whose length is 
comparatively very short, and whose ratio of enlargement is small, this latter loss may 
be expected to be comparatively small and the total loss to be a minimum with a pipe — 
straight or curved — whose angle of divergence — actual or effective — is little greater 
than that giving minimum loss in the diverging portion of the pipe alone. 

Furthermore it would appear possible, although at first sight paradoxical, that in 
a fairly long pipe having a uniform divergence from A x to A 2 at the best angle (about 
11° in a rectangular pipe) the loss of head might even be reduced by reducing the 



THROUGH PIPES OR PASSAGES HAVING DIVERGENT BOUNDARIES. Hi 



angle of divergence still further without altering the length, thus introducing a sudden 
enlargement from A 3 to A 2 and a corresponding loss of head, but at the same time 
reducing the change of velocity, and therefore the loss, in spite of the increasing- 
percentage coefficient, in the taper portion of the pipe. As will be seen from the 
experimental results, given at a later stage in the paper, these conclusions are justified, 
and it becomes possible to design a pipe — often with a considerable reduction in length 
— in which the boundaries are straight, and in which the loss is still appreciably less 
than in a straight taper pipe giving the full enlargement of section with the best 
possible value of 9. 




Fig. 5. 

The total loss of head in such a pipe as shown in fig. 5, and consisting of a straight 
taper pipe terminating in a sudden enlargement of section, is theoretically equal to 

2*7 2g 

where K is obtained from the curves of figs. 2 and 3. As — = — ; while — = — , where 

i\ A 3 vj A, 

A represents the corresponding area this becomes 

2<7 



^Kfl-iAV^YUet 



A 3 A 2 ) j 



or 



2*7 



i d_ A iY + ^< ' ■ 



A 3 / \A 3 



I V }_ 



feet. 



_\A 2 - A x / 

In a rectangular pipe whose breadth increases uniformly from 6 X to 6 3 in a length L 

6 



so that 



b 3 = b } + 2L tan 

'a 

= b 1 + L6 (approximately) where 6 is in angular measure, 
A 3 = A 1 + L0, 



112 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 



and the above expression becomes 



l 0SS = fe^l 2 
2*7 



6 2 -V I V b. + LeJ Kb. + LO bj ) 



The corresponding expression for circular pipes is 



1 0SS = (A^ )2 
2g 






K/l 



»"i+' 



t» » H 



L(?V r„« 



ri+ T; 



• (3) 



(4) 



Differentiating these expressions with respect to 0, the value of and hence of 6 3 or 
r s , giving minimum loss of head, may be obtained. As, however, K varies with 6, the 
resulting expression becomes extremely cumbersome, and the best value is more easily 
obtained by trial of a few values of 9 and the corresponding values of K. Handled 
in this way a solution is readily obtained. 

To test the validity of equations (3) and (4), a series of experiments were carried out 
on rectangular pipes of width 1*329 inch and with an initial breadth of 1*329 inch, 
this increasing uniformly to 2*99 inches, after which the breadth suddenly increased to 
5*315 inches, thus making 63=2*256! and b. 2 = 4b l ; on rectangular pipes of the same 
length and with the same initial and final areas, but with boundaries made to the 
curved templates already indicated ; and on circular pipes having a gradual enlarge- 
ment of section from *50 inch diameter to 1*50 inch diameter, followed by a sudden 
enlargement to 3*0 inches diameter. In the latter pipes the full ratio of enlargement 
was thus 36 : 1. The results were as follow : — 





Length of 

Taper Pipe 

(inches). 


Value of 0. 


Value of 6 
Radians. 


Percentage Loss of Head. 


Experimental. 


Calculated. 


Rectangular 
Pipes. 


9-50 
6-30 
3-10 


10° 
15° 
30° 


•1746 
•2620 
•3492 


192 
23-0 
60-7 


17-6 
23-3 

58-8 


4-70 
3-10 


Curved 


23-4 
39-0 


22-6 
36-3 


Circular ( 
Pipes. j 


3-80 
1*87 


Curved 


21-2 
31-4 


21-1 

32-5 



From these it appears that the experimental and calculated losses are in sufficiently 
close agreement to amply justify the general adoption of the formulae, and these have 
been used to determine the value of 6 giving the minimum loss of head in a number 
of typical cases. The following table shows how the loss in such a pipe compares with 
the loss in a straight taper pipe or in a curved pipe of the same length and of the same 
initial and final sections : — 



THROUGH PIPES OK PASSAGES HAVING DIVERGENT BOUNDARIES. 



113 



1 

<u 
&< 

cz 

'a N 

s 

03 

o 


Ratio of Areas. 


Value of 6 
giving 

Minimum 
Loss. 


Length 

of 

Pipe 

(inches). 


Value of 
6 in 

Straight 

Taper 

Pipe 

of same 

Lengtli. 


Percent- 
age Loss 
of 
Head. 


Loss of Head expressed 

as a Percentage of 

the Loss in 


Straight 
Pipe. 


Curved 
Pipe. 


9 : 1 
f Sections '59 in. x L329 in. ) 
( and 5-315 in. x 1-329 in. | 


10° 20' 
10° 40' 
11° 00 
11° 45' 


17-95 
8-81 
6-49 
4-08 


15° 
30° 
40° 
60° 


16-2 
15-3 
16-2 
20-3 


54-9 
160 
14-2 

... 


81-0 
51-5 


4 : 1 
j 1-329 in. square to ) 
I 1-329 in. x 5-315 in. j 


10° 20' 
11° 00' 
12° 00' 
12° 45' 
15° 30' 


15-15 

11-30 

7-45 

5-48 

3-45 


15° 
20° 
30° 

40° 
60° 


15-6 
15-3 
18-2 
22-0 
29-4 


53-8 
32-6 
22-0 
21-1 


60-5 
53-6 
53-0 


2-25 : 1 
J 1*329 in. square to ^ 
\ 1-329 in. x 2-99 in. / 


10° 40' 
13° 20' 
15° 40' 


6-30 
3-10 
2-29 


15° 
30° 

40° 


15-2 

22-8 
28-2 


51-5 

26-8 
25-2 


430 


Circular Pipes. 


9 : 1 
•59 in. diam. to 1*50 in. diam. 


7° 10' 
11° 00' 


3-80 
1-870 


15° 
30° 


136 
18-2 


523 

27-2 


58-2 
48-0 


4 : 1 

1-50 in. diam. to 3'0 in. diam. 


7° 00' 
13° 30' 


8-575 
2-802 


10° 
30° 


12-7 

20-8 


735 

22-2 


58-5 


2-25 : 1 
2 - in. diam. to 30 in. diam. 


7° 00' 
15° 00' 


5-715 
1-870 


10° 
30° 


13-4 
24-3 


72-5 
35-2 





The loss may be still further reduced by replacing the straight taper portion of 
these pipes by curved pipes having the same initial and final areas and having 
boundaries calculated as already described. With pipes so designed the percentage 
losses become approximately as shown in the table on p. 114: — 

From these results it appears that the effect of replacing the straight taper by a 
curve is proportionately greater in the pipes having the greater length, owing to the 
fact that as the pipe is lengthened the proportion of the total loss which takes place 
during the gradual enlargement is increased, while that taking place at the subsequent 
sudden enlargement is reduced. 

Expressed as a percentage of the loss at a sudden enlargement, the effect of intro- 
ducing the curve is approximately the same for all the pipes, being to reduce this loss 
by about 2'5 per cent, in the rectangular and 1*9 per cent, in the circular pipes. 



§ 7. Summary and Conclusions. 

The following are the chief conclusions to be drawn from the experiments : — 
(a) In a circular pipe with uniformly diverging boundaries the loss of head ex- 
pressed as a percentage of (v 1 — v 2 y/2g varies somewhat with the mean diameter of 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 5). 18 



114 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER 





Area 
Ratio. 


Value of 6 in 

Straight Taper Pipe 

of same Length. 


Percentage 
Loss of Head. 


Rectangular 
Pipes. 


9 : 1 


15° 
30° 
40° 


13-1 
13-0 
13-9 


4 : 1 


15° 
30° 
40° 


12-6 
16-0 
20-0 


2-25 : 1 


15° 
30° 

40° 


12-6 
20-9 
259 


Circular / 
Pipes. 


9 : 1 


15° 
30° 


12-2 
15-8 


4 : 1 


10° 
30° 


11-3 

17-8 


2-25 : 1 


10° 
30° 


12-2 
22-1 



the pipe, and with the ratio of final to initial area, as well as with 9. An increase 
in the mean diameter slightly reduces the percentage loss, as does an increase in the 
ratio of enlargement. For values of 9 between 6° and 35°, however, the differences 
are comparatively small and the loss of head can be expressed by the relationship 



loss = -0110 e^^l-^Uet 
2<7 



where 9 is in degrees. The minimum loss of head is attained with a value of 9 in the 
neighbourhood of 5° 30'. The loss of head at a sudden enlargement of section also 
varies slightly with the smaller diameter d and with the ratio of enlargement m, and 
is given very nearly by the relationship 



loss = 



102-5 + -25m -2-Qd 
100 



{ 2/7 J 



As 9 is increased from 5° 30' the loss rapidly increases to attain a maximum, greater 
in every case than 100 per cent., for a value of 9 in the neighbourhood of 65°. The 
value of 9, which makes the loss equal 100 per cent., varies from 40° to 60°, and in 
practice a sudden enlargement of section is more efficient in the transformation of 
kinetic into pressure energy than is a gradual enlargement in which 9 exceeds this 
critical value. 

(b) By projecting the parallel portion of the pipe into the space bounded by the 



THROUGH PIPES OR PASSAGES HAVING DIVERGENT BOUNDARIES. 115 

diverging walls (fig. 2) the loss is in every case increased, the difference increasing 
with the length of the projection. 

(c) The percentage loss of head in rectangular pipes having one pair of sides parallel 
and the second pair uniformly diverging, varies little with the size of pipe and with 
the ratio of enlargement. 

It can be expressed with fair accuracy for values of between 10° and 35° by the 
relationship 



loss = -0072 6 vili 



,(v 



-^l 2 feet 



2.9 

where is in degrees. This loss is in general greater than in a circular pipe having the 
same value of 6, except where 9 is between 10° and 15°. The minimum loss is obtained 
when 6 is approximately 11°. The maximum loss is apparently obtained with values 
of in the neighbourhood of 70°. 

(d) In pipes of square section the minimum loss of head is obtained with a value of 
in the neighbourhood of 6°, and has a value of about 14'5 per cent. As 6 increases 
the loss becomes much greater — up to 85 per cent, greater — than in the corresponding 
circular pipe. 

(e) By making the pipes trumpet-shaped, with curves designed so as to make 

. y ' = constant, the loss of head in a pipe of given length may be considerably reduced. 
ax 

The proportional saving is greater as the length of pipe is less, and, in the pipes 

examined, varied from 20 per cent, to 60 per cent. 

(/) A still greater saving, combined with great simplicity, may be effected by a 

design giving a gradual uniform enlargement in section from the initial section A x to 

one having an area A 3 , and a sudden enlargement from A 3 to the final section A 2 , as 

shown in fig. 5. The value of 6 in the taper portion of the pipe, which gives a minimum 

loss of head, may be obtained from formulae (3) or (4), p. 112, when the length of this 

pipe is settled. This value varies from 10° to 16° in the rectangular pipes and from 7° 

to 16° in the circular pipes, increasing, for a given length of pipe, as the ratio of 

enlargement is reduced, and, for a given ratio, increasing as the length is reduced. In 

any cases likely to occur in practice its value may be taken as follows, without any very 

great variation from conditions of maximum efficiency : — 



Ratio of enlargement 


9 : 1 


4 : 1 


2-25 : 1 


Value of (rectangular pipe) 


11° 


12° 30' 


13° 30' 


Value of (circular pipe) . 


9° 


10° 30' 


11° 30' 



By this method of construction the loss may be reduced in favourable circumstances 
to about 90 per cent, (in rectangular pipes) and to about 96 per cent, (in circular pipes) 



116 PROFESSOR A. H. GIBSON ON THE RESISTANCE TO FLOW OF WATER, ETC. 

of the minimum possible loss in a uniformly tapering pipe undergoing the full enlarge- 
ment of section. By designing the pipe from A x to A 3 with curved boundaries giving 

\ = constant, the loss may be still further reduced. In the majority of cases 

occurring in practice, however, the additional trouble of calculation and cost of template 
will not be counterbalanced by the slight increase in efficiency which they render 
possible. 

In conclusion, the author would acknowledge his indebtedness to the Government 
Grant Committee, by whose grant in aid the provision of the pipes and apparatus used 
in these experiments has been made possible. 



(117) 



VI. — The Significance of Maximum Specific Electrical Conductivity in Chemistry. 

By Professor John Gibson, Ph.D. 

(Read July 13, 1908. MS. received June 13, 1911. Issued separately October 20, 1911.) 

Photochemical Action. 

The first step made in this investigation was the recognition of increased specific 
electrical conductivity as a general characteristic of photochemical action. It was 
argued that if there be any common characteristic in photochemical changes it must be 
found in the simplest as well as in the more familiar and more complex reactions 
which are characteristic of the metabolism of plants. 

No chemically simpler case suggested itself than the increase in electric conductivity 
of crystalline selenium under the influence of light. * This change might indeed be 
held almost to lie outside the range of chemistry proper, and to belong to the class of 
change often spoken of as merely physical change. But no sharp line can be drawn 
between physical and chemical changes, and the clue proved most useful. 

The action of light on crystallised selenium seems to give evidence of a directive 
action of light rather than of an inherent tendency of the crystalline selenium itself 
towards increased conductivity, for the gain in conductivity persists only as long as the 
exposure to light. Placed in the dark, after exposure to light, crystallised selenium 
reverts to its initially lower conductivity. The same remark applies to all cases of 
photochemical changes which are not permanent. It is otherwise with changes which 
are permanent. A few instances of permanent photochemical change in homogeneous 
systems may be cited : — 

(1) The conversion of yellow phosphorus, under the action of light, into red 

phosphorus. 

(2) The gradual conversion of red amorphous selenium into the black crystalline 

form. 

(3) The conversion of red crystalline mercuric sulphide into the black amorphous 

form. 

These diverse instances of photochemical change are correlated by the fact that the 
change in each case is accompanied by a gain in specific conductivity. 

An apparent exception to this rule is discussed in a very interesting paper by 
Meyer +) on the action of light on the chromo-gelatine film. The main reaction, viz. 
the photochemical reduction of potassium bichromate in presence of soluble organic 

* On Photochemical Action, Proc. Roy. Soc. Edin., 1897, vol. xxi. p. 303. 
t Comp. Zeit. physik. Chemie, lxvi. p. 58. 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 6). 19 



118 PROFESSOR JOHN GIBSON ON 

matter, is undoubtedly associated with increased conductivity, # and the decrease in the 
conductivity of the gelatine film observed is probably due to the heterogeneity result- 
ing from the formation of insoluble gelatine. It is impossible without qualification to 
apply the rule to heterogeneous systems. 

There is an apparent similarity between the action of the short-wave electro- 
magnetic vibrations called " light," in cases such as those above cited, and the action 
of the long-wave electro-magnetic vibrations on the coherers used in wireless telegraphy. 
In this case also there is a resultant increase in the conductivity of the system. 

The Specific Electrolytic Conductivity of Good Electrolytes. 

The line of argument developed in this paper will be greatly facilitated by frequent 
reference to the graphs in Fig. I., which are obtained by plotting as ordinates the specific 
conductivity at 18° C. (K 18 ) of a number of aqueous solutions of electrolytes in 
ohm -1 cm." 1 against the concentration (T) in gram equivalents per kilo of solution 
as abscissae. The advantages of this mode of representing the concentration have been 
discussed in a previous paper. t 

The graph for HC1 may be taken as typical. Starting at the origin with pure 
water, the specific conductivity rises as the concentration of HC1 increases, but 
reaches a maximum (K 18 max = 07646) at a concentration, according to Kohlrausch, of 
(r = 5'0) 18'25 per cent., at 18° C. Beyond this concentration the specific electrolytic 
conductivity of the solutions falls off as the concentration increases. 

The exact concentration corresponding to maximum conductivity is not easily 
determined, as in the neighbourhood of the maximum the conductivity varies but 
slightly with concentration. 

For the sake of brevity, solutions of strong electrolytes will be referred to in what 
follows as being either maximal, ultramaximal, or premaximal, according as their con- 
centrations are equal to, greater, or less than those having maximum specific conductivity. 

Behaviour of Aqueous Solutions containing Hydrogen Chloride. 

Of all known solutions none surpass hydrochloric acid, at comparable concentration, 
in specific conductivity, and none show a higher maximum of specific conductivity. It 
is probable that no other solution has a higher ionic concentration or a higher specific 
conductivity than hydrochloric acid of maximum conductivity. 

The chemical properties of hydrochloric acid vary in a remarkable manner with the 
concentration. In dilute premaximal solutions hydrochloric acid behaves as a very 
stable acid. In concentrated ultramaximal solutions it is readily oxidised and acts as 
a reducing agent. In dilute premaximal solutions it favours hydrolysis. In highly 

* This is clearly recognised by Meyer, loc. cit. 

t J. GiiiSON, Trans. Roy. Hoc. Ediu., 1905-6, vol. xli. p. 241. 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 



119 



Fig. I. 



0/234 


5 ff 7 S 


3 


/0 


11 


12. 


13 


/4 


15 


/6 


/7 


18 O 




r 








r 


4 


i 


s 


































































"^Z^N 






























































\ M -n 




























































(\\J 


"^ ^ 




























































jIJ-X 






























































7<3!£jy 


^l!l £ 






























































4Attaz 4^_ 


























































































































11 


u . C j4iCS 




























































if 






























































1 1 j 






























































-iM I 






























































-iM -i- 






























































*W 1 






























































flt 4- 


_C I J 




























































flt t 


-4 4 




























































tt t- 


_I H 




























































4J - 


4 H 




























































44 3 - 


-4 1- 




























































41 4 - 


-4 4 




\\ 
























































-it 4 - 


_ H ^ I 




v 
























































tt t - 






\ 


\ 






















































45-£ - 


n J*£ zW- 




^ 


\ 


H 


f 


















































tl-, / 


- v t- 






\ 






















































IV / 


- 3 t 






\ 






















































t 4 t- 


- v- 4 






\ 






















































£p i l- 


4 4 








\ 




















































*4 14 - 


> ^ 








\ 




















































t 4 - 


C r- 








\ 




















































4 t - 


4 t 








\ 


* 


















































t-A ' - 


J 4 










£ 


















































JL _ 


\-4 










\? 


~L 
















































4444 - 


44 










\ 


-if* 
















































4 tt - 


T H 










) 


















































4 tt - 


4 ■- 












\ 
















































I&t - 


rctf 1 












\ 
















































it-. 


^72 


f/* 










\ 
















































b£. 














1 
















































It -4 /> 
















\ 














































M^ - 
















\ 














































Me- 
















\ 














































lt 






























































tt ^ 


















\ 






























































\ 




































r 








7 ~f 


















\ 




































I 








1 T_j 


"A&Dir 
















\ 




































/ 








J £4 


















\ 


































t? 










^34f£ : 




















\ 






























1 


r 






























\ 
































I 










w\\\ 




















\ 
































J 










jE|/T 




















\ 
































f 










tpi/l 


./* S.K." 




















\ 








































1 1 






















\ 




























I 












O * 


£i(*-Ws- _ 




















\ 




























/ 












Iff ^ 






















\ 




























/ 












Si -,£ 


















































/ 








ff 


1 


St #-- 
























































/ 


Va 


w 


3 ? 






















































/ 








I1/1 / ^" 


zCi^. "^ 














































/ 






/ 
























































/ 




/ 










4j7 1 ~7 * 


-*"^£^A 














































/ 


/ 


i 




»Va 


N 


* 


4tt j_j ' 


















































/ 




/ 








III ( / 


\ .v YB%'"i 












































/ 




/ 












Ml ////''*" 


N •" \ j 












































/ 


/ 














1 ' L.Lm&'~~^S 


"S V 












































/ 


/ 














II i / K/ j*'~Si_ 


N v \ v. 












































/ 
















4L.JUV €& 








































* 




?a 


7 


/ 














"I" £7 


e-^5 ^v ^ 




































•? 






% 


// 






















ii 


CI 




































\ 


// 


j 


•N 


*? 


b'6 


u 






If 17 4? 


>. ^Mo?l 


; 








































1 




/ 














£3 












































i t 


/ 
















£ 


k — 










































4 


' 
















1 ^^ — *~^^ 


f 7 1 / ^ ' W . 










































/ 










H 


5 


v* 




<JL ^ 










































4 


1 













































































































































































































6 

5 

« 

J 

2 

/ 
7000 

9 

8 

7 

6 
6500 

4 

J 

2 

/ 
6000 

9 

8 

7 

6 
SBOO 

3 
2 
/ 
5000 
9 

8 
7 
6 

#5 00 
4 

2 
/ 
4000 
9 
8 

7 

6 
5500 

4 

3 

Z 

I 
3000 

9 

8 

7 

6 
Z500 

4 

3 

2 

/ 
2000 

9 

8 

7 

6 
1500 

4 

3 

2 

/ 
1000 

900 

800 

700 <*> 

600 ^ 

SOO * 
400 <^ 
300 X 
ZOO 

100 



\ 



O I z 
V s> 



to II IZ 13 14 IS 16 17 IB 



r= 



gram equivalents per kilogram =m-^sp. gr.= P er cent ; xl ° . The values for 10 4 K ]8 are taken from Kohlkausch 

gram equivalent 



ami Holbtjrn, Leitvermogen der Electrolyte, except those for HBr and HI, which were determined in this laboratorv 



120 PROFESSOR JOHN GIBSON ON 

concentrated ultramaximal solutions it favours abstraction of water and changes which 
are accompanied by the formation of water, such as esterification. These different pro- 
perties may all be summed up by simply stating that solutions of hydrochloric acid 
tend to gain in specific conductivity. 

Saturated solutions of certain chlorides may be used as indicators of this tendency. 
If, for example, a small quantity of a solution of hydrogen chloride be added to a 
saturated solution of common salt containing a few minute crystals of salt in suspension, 
either some of the undissolved salt will dissolve, or else some salt will be thrown out of 
solution. It was found that precipitation occurs only when the hydrochloric acid added 
is in ultramaximal solution. The same condition holds for precipitation by hydrochloric 
acid of the chlorides of potassium, ammonium, and rubidium from their saturated 
solutions.* Premaximal solutions of hydrochloric acid may be distinguished in this 
way from ultramaximal solutions even when their concentration differs from the 
maximal solution by only ±1 per cent. HC1. Thus, precipitation was not observed 
with 17 per cent. HC1 or less, and invariably observed with 19 per cent, and more. 



The Behaviour of Aqueous Solutions of Hydrogen Chloride towards Dissolved 
Oxygen and Dissolved Chlorine respectively. 

The behaviour of solutions of hydrogen chloride towards dissolved chlorine on the 
one hand, and towards dissolved oxygen on the other, presents another case in point. 

Chlorine decomposes water, especially under the influence of light. The action 
is partially illustrated by the equation 

2H 2 + 2Cl 2 ^4HCl + 2 . 

This reaction is a reversible one, or, in other words, under certain conditions dissolved 
oxygen and hydrogen chloride react so as to produce free chlorine and water. Current 
theories do not enable us to predict the concentration at which equilibrium would be 
established. Assuming a tendency towards increased conductivity, the progress of the 
reaction in either direction should be associated with a gain in specific conductivity. It 
is owing to this tendency towards increased specific conductivity that concentrated ultra- 
maximal solutions of hydrogen chloride, after long keeping in partially filled clear glass 
bottles, contain free chlorine, due to the oxidation of the hydrogen chloride. Numerous 
solutions were examined, but free chlorine was never found to persist in solutions which 
had been diluted to below 18 percent. HC1, i.e. to below the concentration corresponding 
to maximum specific conductivity. So long as the solution is premaximal, free chlorine 
disappears from it, with formation of hydrogen chloride and free oxygen, the 
solution thereby gaining in specific conductivity. On the other hand, when the hydro- 
chloric acid is in ultramaximal solution, hydrogen chloride is oxidised by the oxygen 

* (Jihson and Denison, Proc. Roy. Soc. Edin., vol. xxx. p. 562, 1909-10. 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 121 

of the air into water and free chlorine, the solution here also gaining in specific 
conductivity.* 

The equation for this reaction may be put in the following form — 

(x + 4)HC1 + yR 2 + 2 — xHCl + (y + 2)H 2 + 2C1 2 . 

In this form the equation clearly indicates that the progress of the reaction in the 
direction indicated by reading from left to right involves a dilution, whereas its 
progress in the opposite direction involves a concentration of the hydrochloric acid. 
When the concentration of the free chlorine or of the dissolved oxygen is so small 
as to have only a negligible influence on the conductivity, the concentration of the 
maximal solution of hydrogen chloride coincides with the concentration at which 
equilibrium is established. 

Oxidation of Hydrogen Chloride in Aqueous Solution by Chromic Acid. 

Experiments were made in order to determine the relationship between the 
specific conductivities of known solutions of hydrochloric acid of different con- 
centrations and the time required to oxidise a small constant proportion of the 
hydrogen chloride contained in these solutions. 

Chromic acid was chosen as the oxidising agent for this purpose, because it gives 
rise to a marked change of colour on reduction. A difficulty arises, however, owing to 
the fact that the addition of this oxidising agent, itself a strong electrolyte, affects the 
conductivity of the whole system. Practically nothing is known of the conductivity 
of such mixtures. 

To obviate this difficulty, the quantity of chromic anhydride added was small. 
Further, the ratio of Cr0 3 to HC1 was kept the same for all the solutions. Under these 
conditions no serious error is incurred by regarding the conductivity in each case as 
being simply that of the aqueous solution of hydrogen chloride. It was found 
that the oxidation of the hydrogen chloride, as indicated by the change of colour 
due to the reduction of the chromic acid, proceeded more and more slowly the 
more closely the concentration of the solutions of hydrogen chloride taken approached 
that corresponding to maximum specific conductivity. 

It was not found possible to determine the end points with accuracy in the 
diluter solutions, as in them the reaction is very slow. The results of one series of 
experiments are given in Table I. 

* Comp. Backelandt, Bull, de I'Acade'mie royale de Belgique, 3 me serie, t. xl., N r 3, 1886. 



122 



PROFESSOR JOHN GIBSON ON 



Table I. 
Hydrochloric Acid and Chromic Acid. 







V = Kmax. - K- 






Per cent. 
HCl. 


r. 




Time in 
days. 


Time in 
months. 








y x 10 2 . 






32 


878 


137 


0-25 


•008 


30 


8-23 


101 


0-3 


•01 


28 


7-68 


73 


G 


•2 


24 


6-58 


25 


38 


1-3 


21-7 


5-95. 


8 


118 


3-9 


20 


549 


1 


174 


5-9 


1822 


5 00 





>700 


25 circa 



N.B. — With 16 per cent, and 12 per cent. HCl the reduction was not 
nearly complete after four years' standing. 

In the graph (Fig. II.) the times required for complete reduction of the Cr0 3 are 
plotted as abscissae against the specific conductivities of the respective solutions as 
ordinates. The maximum conductivity of hydrochloric acid is taken as origin, so that 
the ordinates indicate decrements of conductivity from the maximum, that is to say, 
they indicate the values for y = K max . — K. 



>f\ 



i4 * 



Fig. II. 



2 3 4 5 6 7 S 9 10 II IZ IZ 14 IS 16 17 18 19 20 2/ ZZ ZS Zt 25 26 27 28 

Time in months — *" 



From this series of experiments and from a number of other cases it appears that 
the tendency towards increased conductivity is greater the greater the possible gain in 
specific conductivity ; or, in other words, the tendency increases with the value of 
y = Kmax. - K. There is some reason to think that it may also increase with the value 
for K, n . r . 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 123 

Hydrochloric Acid as an Esterifying Agent. 

Consider a reaction such as that between oxalic acid and ethyl alcohol mixed in 
equivalent proportions 

2C 2 H 5 OH + H 2 C 2 4 = (C 2 H 5 ) 2 C 2 4 . 

The reaction from left to right is soon counterbalanced by the alteration in the 
concentration of the products, and equilibrium is reached long before the whole of the 
alcohol has been esterified. By constant withdrawal of the water formed during the 
process the esterification may approach completion. But the same end can be attained 
in another way, without the removal of the water, viz. by passing HC1 gas into the 
mixture. By doing this, the chemist substitutes a system in which, as a result of the 
progress of the reaction from left to right, a great gain in specific conductivity is 
possible, for a system in which little or no such gain is possible. The tendency towards 
maximum specific conductivity is brought into play, and becomes the predominating 
factor, and the velocity of the reaction from left to right is therefore much greater. 
The possible gain in specific conductivity (y = K max . - K) of the original mixture of 
alcohol and oxalic acid is very small, while that for the mixture containing HC1 in 
addition is much greater. So long as the hydrochloric acid solution is ultramaximal, 
the dilution due to formation of water implies a gain in conductivity, but in premaximal 
solutions the case is reversed. Thus : — 

(a) In dilute premaximal solutions hydrochloric acid accelerates catalytically the 

hydrolysis of esters. 

(b) In concentrated ultramaximal solutions it favours the esterification of Alcohols 

by weak acids, instead of itself forming an ester. 

In (a) there is a gain in conductivity due to the increase of the concentration of 
hydrogen ions. In (b) there is a gain in conductivity due to the decrease in the 
number of molecules which, being mostly associated and not split up into free ions, e.g. 
(H 2 C 2 4 ) and (C. 2 H 6 0), lower the conductivity of the strong acid, while at the same 
time the water formed by the reaction greatly increases the specific conductivity, by 
diluting the highly concentrated ultramaximal acid solution. Were the strong acid 
itself to form an ester, there would not be this great increase in conductivity. 

a;H 2 C 2 4 + 2zC 2 H 5 OH + 2/HC1 ^ a(C 2 H 5 ) 2 C 2 4 + ? /HCl + 2a;H 2 0. 

Action of Hydrogen Chloride on Ethyl Aldehyde, Aldol, and 
Crotonic Aldehyde. 

When acetaldehyde is treated with hydrochloric acid of the concentration of 
maximum conductivity, or with a rather more dilute solution, aldol is formed. 



124 PROFESSOR JOHN GIBSON ON 

In this case water is neither abstracted from nor added to the aldehyde. The 
assumption of a tendency towards increased conductivity suggests that there should be 
an increase in specific conductivity accompanying the reaction. How can this be so \ 

Aldehyde is a non-electrolyte, and, when added to a solution of hydrochloric acid, 
it lowers the specific conductivity of the acid solution. This lowering of the specific 
conductivity is greater the greater the number of molecules of the non-electrolyte 
which are added. AVhen molecules of aldehyde unite to form molecules of aldol, the 
number of molecules of non-electrolyte is halved, and the specific conductivity must, 
therefore, be increased. Further, so long as the reaction does not involve dehydration, 
the possible gain in specific conductivity thus brought about will probably be at a maxi- 
mum when the acid has the concentration corresponding to maximum specific conductivity. 

In order to demonstrate that a rise of specific conductivity does actually accompany 
this reaction, the following experiment was tried : — 

The specific conductivity of a mixture of aldehyde and maximum conductivity 
hydrochloric acid (18 per cent, to 19 per cent.) was measured from time to time, the 
mixture being kept in a thermostat at 18° C. During the three days subsequent 
to the preparation of the mixed solution its conductivity rose steadily, and by 
the third day had risen 12*5 per cent., or from 10 3 K 18 = 203 to 10 3 K 18 = 228. 

When aldol is treated with a solution of hydrochloric acid somewhat stronger than 
that having maximum specific conductivity, that is, with a slightly ultramaximal 
solution, it loses water and is converted into crotonic aldehyde. 

On the other hand, crotonic aldehyde in presence of a much more dilute hydro- 
chloric acid takes up water, re-forming aldol. 

The aldol condensation, the dehydration of aldol, and the hydrolysis of crotonic 
aldehyde all take place in hydrochloric acid solutions, but the conditions in each case 
would appear to be such that an increase of specific conductivity accompanies the 
progress of the reaction. 

Thus the tendency towards increased specific conductivity may show itself 
otherwise than by favouring dehydration or hydrolysis. The effect of adding 
molecules of a non-electrolyte to a conducting solution is in general to decrease 
the specific conductivity of the solution. This lowering of the specific conductivity 
depends primarily on the number of molecules of non-electrolyte added, so that a not 
inconsiderable increase in the specific conductivity of the solution may result from the 
polymerisation and consequent decrease in the number of such molecules. The 
polymerisation of molecules of the non-electrolyte present in a conducting solution may 
be regarded as favoured by the gain in conductivity which results from the removal or 
decrease in the number of such molecules. 

Solutions which can be obtained by adding non-electrolytes, or relatively weak 
electrolytes, to solutions of strong electrolytes, may be conveniently distinguished 
by the prefix " sub." A solution obtained by adding a non-electrolyte to an ultra- 
maximal solution may, on dilution, gain in conductivity, and may be converted into a 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 125 

submaximal solution, but it cannot be converted into a maximal solution by merely 
adding water. It is thus sub-ultramaximal. 

Similarly, a solution may be sub-premaximal, in which case it will gain in con- 
ductivity on becoming more concentrated by loss of water, but cannot become a 
maximal solution by mere increase of concentration. 

All such solutions may be expected to gain in conductivity as a result of the 
polymerisation * of the molecules of non-electrolytes or weak electrolytes which they 
contain. 

The greater the concentration of the non-electrolyte present, the greater is the 
possible gain in conductivity due to its polymerisation, and accordingly the greater is 
the tendency, ceeteris paribus, for such polymerisation to take place. The tendency 
towards polymerisation is not confined to solutions having a concentration of the 
electrolyte on one side only of maximum conductivity, but clearly belongs to sub- 
premaximal solutions as well as to sub-ultramaximal solutions. 

In the synthesis of organic compounds, polymerisation or condensation without 
dehydration is often brought about by means of strong electrolytes. It is probable 
that these reagents will be found, as a rule, most effective for this purpose in the form 
of premaximal solutions. Where condensation or polymerisation accompanied by 
dehydration is desired, ultra-maximal solutions will prove, generally speaking, more 
effective. 

Hydrochloric Acid and Cobalt Chloride. 

Concentrated solutions of hydrogen chloride change the pink colour of hydrated 
cobalt chloride to a blue-green. A similar colour change is produced if the crystallised 
salt be heated so as to drive off water of crystallisation. Granting that the change of 
colour in the first case is also accompanied by dehydration, the assumption of a tendency 
towards increased specific conductivity suggests that premaximal solutions of hydro- 
chloric acid should not cause the change of colour, because such solutions lose in 
conductivity by dilution. To test this, 50 c.c. each of a series of solutions of HC1 of 
known concentration were severally mixed with two drops of a concentrated aqueous 
solution of cobalt chloride, and then arranged side by side, in order of concentration of 
HC1. It was then seen that in the weaker solutions up to 16 per cent. HC1, the pink 
colour was not altered. A solution of 18*2 per cent. HC1 showed only a faint tinge of 
purple, but once past this latter concentration the change in colour was more and more 
marked. Using cobalt chloride as an indicator, it is quite easy in this way to make up 
a solution containing very nearly 19 per cent. HC1 without any measurement or 
weighing. 

Solutions of hydrobromic acid showed a similar behaviour with cobalt chloride. 

* The term "polymerisation" is here intended to include cases of condensation of the aldol type which are not 
accompanied by a permanent hydrolysis or dehydration. 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 6). 20 



126 PROFESSOE JOHN GIBSON ON 

No change of colour was observed with solutions weaker than 34 per cent. HBr, which 
is the concentration of maximal hydrobromic acid * (K max = 07590 ; r„, ax = 4*24). 



Behaviour of Aqueous Solutions of Hydrogen Iodide. 

In Aqueous solutions of hydrogen iodide the tendency towards increased specific 
conductivity is frequently masked, owing to the strong affinity between dissolved 
oxygen and the hydrogen of hydrogen iodide. 

Aqueous solutions of hydriodic acid do, however, exhibit the tendency to gain 
in specific conductivity. 

A well-known method for the preparation of hydrogen iodide is to suspend iodine in 
water and pass in hydrogen sulphide. This reaction, however, cannot be used to obtain 
solutions of pure hydriodic acid of a higher concentration than that of the maximal 
solution which has a conductivity at 18° C. of K raax = 740, r = 34, # for in ultra- 
maximal solutions of hydrogen iodide an increase in the concentration of hydrogen 
iodide involves a decrease in specific conductivity. So long as the acid is premaximal 
the dissolved iodine is completely converted into hydrogen iodide, and the solution 
becomes colourless. Ultramaximal solutions remain coloured, however vigorous or 
long-continued the current of hydrogen sulphide may be. 

In all these systems where the strong affinities are nearly balanced, the tendency 
towards a gain of specific conductivity becomes effective and determines the position 
of equilibrium. 

The tendency of chlorine to unite with the hydrogen of water is of the same order 
as that of oxygen to unite with the hydrogen of hydrogen chloride, and similarly the 
tendency of iodine to unite with the hydrogen of hydrogen sulphide is of the same 
order as that of sulphur to unite with the hydrogen of hydrogen iodide. 

The tendency towards increased conductivity is masked, in an aqueous solution 
containing hydrogen iodide and oxygen, because the affinity of the hydrogen of 
hydrogen iodide for oxygen is enormously greater than that of the free iodine for the 
hydrogen of water. 

The addition of free iodine to a solution of hydrogen iodide lowers the conductivity 
of the solution. The resulting solution can decompose hydrogen sulphide only so long 
as there is a consequent gain in conductivity. 

In an ultramaximal solution of hydriodic acid the tendency towards increased 
specific conductivity may even induce the action of hydrogen iodide on sulphur, with 
production of hydrogen sulphide, free iodine, and probably HI 3 . 

mH 2 + (n + 2)HI + S ^ mH 2 + nKI + II 2 S + I 2 
mll 2 + (n + 3)TII + S — mH 2 + nHI + IT 2 S + HI 3 . 

* According to conductivity determinations made in this laboratory. 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 127 

If the quantities of free iodine and hydrogen sulphide be kept small, the direction 
of the whole reaction appears to depend mainly, if not solely, on whether the solution 
of hydriodic acid be premaximal or ultramaximal, and the velocity of the reaction in 
either direction depends on the value for y = K max . - K. It is only from premaximal 
solutions of hydrogen iodide that the last traces of free iodine disappear altogether. 



Behaviour of Nitric Acid. 

Nitric acid and highly concentrated aqueous solutions of nitric acid undergo partial 
decomposition when exposed to light. The concentration of nitric acid corresponding 
to maximum conductivity lies between 28*5 and 29 per cent. HN0 3 . At lower con- 
centrations nitric acid is quite stable under the influence of light. 

Peroxide of nitrogen and water are products of this photochemical decomposition. 
As this decomposition progresses there must be in such concentrated solutions a rise in 
conductivity, consequent upon the dilution of the acid. 

In order to throw light on this behaviour of nitric acid, a number of solutions were 
prepared having known concentrations higher and lower than the maximum. Pure 
nitric oxide gas was passed into these solutions. With the solutions having concentra- 
tions higher than nitric acid of maximum conductivity colours were obtained passing 
from yellow-brown in the more concentrated solutions through green in less concentrated 
solutions, to pure blue in solutions only slightly more concentrated than the maximum 
acid. None of the solutions weaker than the maximum acid had any oxidising action 
on the nitric oxide, and they therefore remained perfectly colourless. Thus in nitric 
acid of 36 per cent, the blue colour was readily obtained, whereas an acid of 28 
per cent. HN0 3 remained quite colourless. Decomposition of nitric acid in pre- 
maximal solutions could only imply a lowering in specific conductivity. They are not 
decomposed either by light or by nitric oxide. 

The Action of Sulphuric Acid on Cane-Sugar. 

According to its concentration sulphuric acid may be used either to hydrolyse or to 
dehydrate cane-sugar. 

If to excess of concentrated sulphuric acid containing 84 per cent. H 2 S0 4 a little 
of a concentrated solution of cane-sugar be added, charring sets in rapidly and is soon 
complete. On progressive dilution from this point onwards the time taken to char the 
sugar increases. When the acid has been diluted to about 30 per cent. H 2 S0 4 , it may 
be left for many days along with dissolved cane-sugar at the ordinary temperature, and 
the solution may even be boiled for several minutes without the slightest indication 
of charring. 

Finally, in solutions of more dilute sulphuric acid, the chief reaction is of an 



128 



PROFESSOR JOHN GIBSON ON 



opposite nature, for, instead of being dehydrated, the cane-sugar takes up water 
from the solution and is hydrolysed, forming ultimately a mixture of glucose and 
fructose. 

84 per cent, sulphuric acid consists wholly, or almost wholly, of the monohydrate 
IL>S0 4 ,H 2 0, and, according to the general rule that single substances are very poor 
conductors, there is a minimum of specific conductivity at this concentration 
(K = 0'0979). From this point onwards a rapid increase in conductivity accompanies 
progressive dilution until a concentration of about 30 per cent. H 2 S0 4 is reached, 
which is the concentration of the maximal acid. 

K max . at 18° C. = 07388. At 84 per cent. H 2 S0 4 the value for ?/ = K niax .-K is 
07388 -0-0979 = 0-6409. 

As the acid is progressively diluted from this point onwards the value for y falls, 
and with it the tendency of the acid solution to dilute itself by dehydrating the sugar 
decreases. When the maximum conductivity is reached, the tendency towards dilution 
disappears and the solution is relatively inert towards the sugar. At concentrations 
less than that corresponding to maximum conductivity the solution tends to concentrate 
itself by giving up water to the cane-sugar. Table II. and the corresponding graphs 
on Fig. III. show the results of two series of experiments made with solutions of sul- 
phuric acid containing varying proportions of cane-sugar. 

The times are those which elapsed between the date of mixing and the appearance 
of a brown tint, indicative of incipient charring. 



Table II. 
Cane-Sugar and Sulphuric Acid. 











A. 


B. 


Per cent. H 2 S0 4 . 


r. 


10 3 K. 


io 3 y ls . 


Time in 
Months. 


Time in 
Months. 


60 


12-23 


373 


366 




■05 


50 


10-20 


541 


198 




•37 


45 


9-18 


616 


123 


•06 


■ • • 


40 


8-16 


680 


59 


•40 


1-1 


36 


734 


715 


24 


1-6 


2-9 


33 


6-73 


733 


6 


5-6 


7-7 


30-5 


6-22 


739 





11-9 


12-4 


26 


5-30 


722 




uncertain 












(very long) 





A refers to solutions containing 1 gram sugar to 50 c.c. acid. 
B „ „ 0-3 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 



129 



When, as in Fig. III., the conductivities of the solution of the pure electrolyte are 
plotted instead of the conductivities of the actual solution (in this case sulphuric acid 
and sugar), then the influence of the presence of the non-electrolyte molecules shows 
itself by a distortion of the graph, such that it cuts the time axis. The greater the 
concentration of the non-electrolyte, the greater the distortion, and the greater will 





40 




30 




20 




10 




300 




90 




80 




70 




60 




50 




40 




30 




20 




10 




ZOO 




90 




80 




70 




60 




50 




40 




30 


X 

>•> 


20 

10 

100 

90 


2 


80 
70 
60 


CO 

o 

1—1 


50 
40 




30 




20 




10 








Fig. III. 



f 


I 




■ 


— ■ 


















































































































































































































































































































































































































































































































































































































































































































































1 




















































' 




































































































































































































































































1 


















































I 


u 




























































































































































\ 




















































\ 




















































\ 
































































































































































































































































































































































































































1 






> 


• 


• 


4 




/ 




t 






r 


6 


1 


t 


> 


1 





1 


1 


1 


2 





Time in months 



be the inclination of the graph towards the time axis. In Fig. III. two graphs for 
solutions of sulphuric acid containing respectively : 

(A) 1 gram cane-sugar per 50 c.c. of solution, 

(B) 0'3 gram cane-sugar per 50 c.c. of solution, 

are given. The graphs show clearly the effect of an increase in the concentration of 
the non-electrolyte. 

In A the proportion of cane-sugar is sufficient to cause a marked distortion of the 
graph. 

In B the distortion is much less, as the conductivities are comparatively little 
different from those of pure solutions of sulphuric acid. 



130 PROFESSOR JOHN GIBSON ON 



Sulphuric Acid and Formic Acid. 



The action of sulphuric acid on formic acid affords a striking example of the 
relationship between the value for ij and the dehydrating power of sulphuric acid. 
Ultramaximal solutions of sulphuric acid of such concentration, that the value for y is 
considerable, readily effect the dehydration of formic acid, with production of carbonic 
oxide ; but in solutions where the value for y is small this dehydrating power is also 
small, and ultimately disappears on progressive dilution as the value for y approaches 
zero, that is, as the concentration of the ultramaximal solution approaches that of the 
maximal solution, i.e. 30 per cent. H 2 S0 4 . 

The following experiments were tried : In each experiment a small quantity, 
5 c.c, in volume of anhydrous formic acid was mixed with 100 c.c. of one of a series of 
standard solutions of sulphuric acid having the concentrations given in Table II. 
The mixture was heated in a distilling flask, and when necessary a slow current of 
carbonic anhydride was used to sweep out the last portion of carbonic oxide evolved, 
this gas being collected and measured over a strong solution of caustic soda. With 
the more highly concentrated acid solutions carbonic oxide was given off freely, but 
when the concentration fell to r = 8 # l. (y = - 123 ohm" 1 cm. -1 ), only a few cubic 
centimetres of carbonic oxide were given off. With T = 7'S (y = 0'025), still less, 
and with r = 6 # l, or lower, no carbonic oxide was obtained. (See Fig. I.) Evidently 
the sulphuric acid does not dehydrate formic acid in premaximal solutions. 

Summary. 

The examples discussed so far point to a remarkable relationship between the 
velocity of very many reactions and one particular quality of the media in which they 
occur, viz. their tendency towards increase of specific conductivity, this tendency 
being measured by the value for y = K max . — K. 

These examples may be summarised and correlated in the following manner : — 

Homogeneous chemical systems which undergo change either of themselves or under 
the influence of the electro-magnetic vibrations which we call " light" change so that 
their specific electrical conductivity is increased, unless when coerced in an opposite 
direction by stronger chemical affinities. 

The justification for such an hypothesis must lie in its usefulness. It must make it 
possible to predict correctly the results of hitherto untried experiments which it 
suggests, and lead to the correlation of hitherto uncorrelated phenomena. The proof 
offered in this paper is a cumulative proof. It is drawn from a great variety of instances 
widely different in character. 

Apparent exceptions are capable of classification and correlation. 

For instance, owing to strong chemical affinities, it is impossible to prepare a solution 
containing any considerable concentration of hydrogen ions along with a corresponding 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 131 

concentration of hydroxyl ions. A solution containing a concentration of hydrogen 
ions such as is found in hydrochloric acid of maximum specific conductivity (circa 18 
per cent. HC1), along with a concentration of hydroxyl ions such as is found in aqueous 
caustic potash of maximum specific conductivity {circa 27 '5 per cent. KOH), would be a far 
better conductor than any known solution, as these two ions have greater mobility than 
any others ; but the strong affinity which determines their immediate association and 
the formation of water brings about a solution having a specific conductivity lower than 
that of either solution taken separately. Thus, on mixing the best conducting acid 
solution known with an equivalent quantity, that is, with a nearly equal volume, of the 
best conducting alkaline solution known, i.e. maximal caustic potash, the tendency for 
the hydrogen ions of the acid and the hydroxyl ions of the alkali to associate overcomes 
and masks the tendency towards an increase in specific conductivity, so that only the 
much less mobile potassium and chlorine ions are left as the chief carriers of electricity 
in the neutral and less highly conducting solution of potassium chloride. 

The hypothesis is applicable to homogeneous, i.e. single-phase systems. It cannot 
even be formulated for heterogeneous systems, since the term "specific conductivity" 
applied to a heterogeneous system has no meaning. There are, however, many cases 
where it is possible to apply the hypothesis usefully, and to predict the course of events 
correctly, although the system is, or becomes, heterogeneous. Thus, in cases where a 
rearrangement resulting in alteration of specific conductivity brings about the separa- 
tion of a non-electrolyte from a highly conducting solution, the system no doubt becomes 
heterogeneous, but its heterogeneity may be disregarded whenever the actual change of 
conductivity would not have been materially affected had the non-electrolyte remained 
in supersaturated solution. Heterogeneity confuses the issue only when marked changes 
in conductivity are the direct result of the appearance of the new phase or phases. 

The separation of the non-electrolyte sulphur, in the action of hydrogen sulphide on a 
solution of iodine in hydriodic acid, is a case where heterogeneity may be disregarded. 

The precipitation of barium sulphate and silver chloride, on mixing equivalent 
solutions of silver sulphate and barium chloride, is a case where the hypothesis is 
clearly not applicable. In this and similar cases of the double decomposition of salts 
the removal of electrolytes from the solutions necessarily implies a lowering of its 
conductivity, while at the same time the system becomes heterogeneous. 

Application of the Hypothesis to Plant Chemistry. 

This investigation was originally undertaken with the special object of throwing 
further light upon the chemistry of plant metabolism. With the exception of the first- 
cited instances of photochemical action, the examples and reactions discussed so far 
have all been reactions in which strong mineral acids play an essential part. This is 
not a matter of choice, the reason being that the tendency towards increased con- 
ductivity depends on the value for y = K max . - K, which is necessarily small when 



132 PROFESSOR JOHN GIBSON ON 

&max. is small. It is therefore in solutions where K„ lllx . is greatest, that is, in concentrated 

solutions of the strongest acids, that the clearest evidence of the existence of the 

tendency towards increased conductivity is to be found. At first sight this would seem 

to lead far away from the chemistry of plant life, where mineral acids, and more 

particularly concentrated solutions of free mineral acids, are characteristically absent. 

When the strong chemical affinities are not balanced, but directly brought into play 

as in the action of acids on bases, and generally wherever the strong tendency of the 

+ 
charged ions H and OH to neutralise each other predominates, there the weaker 

tendency towards increased conductivity is masked. Although the systems hitherto 
discussed comprised strong mineral acids, strong chemical affinities were uniformly 
more or less completely counterbalanced. Thus the examples discussed have com- 
prised, among others, the action of nitric acid on nitric oxide ; the action of 
hydrochloric acid on chromic acid and on aldehyde ; the action of iodine on hydrogen 
sulphide, and the action of sulphuric acid on formic acid and on cane-sugar. 
Chemical systems in which the tendency towards increased conductivity is very marked 
and clearly recognisable resemble the chemical systems characteristic of plant life in 
this, that they are conducting systems containing good electrolytes, but with the strong 
affinities in abeyance. In plant chemistry the place of strong acids is taken by salts 
derived from the soil or, in the case of marine plants, from sea water. There is this 
further resemblance, that the solutions of many of the salts specially useful to plants, 
show maxima of specific conductivity, as do the strong acids. The conductivity of 
solutions of such salts have a specific conductivity of the same order as those of 
the strong mineral acids of corresponding concentration, their conductivities ranging in 
general from about one-third to one-sixth of those of the strong acids. So far as the 
insufficient data permit, it would appear that by merely selecting those salts which 
give the best conducting solutions and exhibit maxima of specific conductivity, we 
obtain an indication of the kind of salts most generally useful in plant chemistry. 
From the salts whose conductivities are given by Kohlrausch and Holburn, 
calcium, magnesium, lithium, and manganese chlorides, potassium carbonate, fluoride 
and acetate, are thus singled out. (See Fig. I.) 

Reactions characteristic of plant chemistry are generally represented by equations 
which do not include good electrolytes. The fact that in plant metabolism the presence 
of good electrolytes, i.e. metallic salts, is indispensable has not hitherto been explained. 
The suggestion is here made that the tendency towards increased specific conductivity is 
an essential and determining factor in plant chemistry. It is not easy to obtain experi- 
mental evidence in favour of this suggestion in respect of foliage leaves, owing to the 
structural complexity of the tissues in which the reactions take place. The chemical 
reactions occurring in plants must be associated with changes in the local concentration 
of the sap solutions, which constitute the medium in which they occur, and consequently 
they must lie associated with local changes in specific conductivity. 

There is in foliage leaves a striking periodic alternation between dehydration and 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 133 

hydrolysis, namely, the diurnal change from sugar to starch, and again from starch to 
sugar. # Foliage leaves favour evaporation and sap concentration during insolation, by 
exposing a great surface, and by opening their stomata. They not less surely favour 
subsequent sap dilution by a checked evaporation, due to closure of the stomata when 
insolation ceases. Thus sunlight, evaporation, sap concentration, and conversion of 
sugar into starch, alternate with darkness, checked evaporation by closure of stomata, 
dilution with fresh sap, and consequent reversal of the reaction, and reconversion of 
starch into sugar. To increase the concentration of a premaximal solution, and to dilute 
an ultramaximal solution, is in both cases to bring about an increased conductivity. 
Bearing this in mind, we are able to correlate these two apparently opposed reactions 
as both tending towards increased conductivity. 

Local sap concentration must surely occur in the intracellular regions of the meso- 
phile where the sunlight is concentrated by the cell lenses of the epidermis on the 
light-absorbing chloroplasts congregated together close to the opened guard cells, which 
permit a free escape of water vapour. It is here that metabolism is most active and it 
is very probably here that sugar is chiefly formed. 

Applying the hypothesis, it follows that when the sap concentration increases, 
and the solution becomes ultramaximal, the sap should tend to dilute itself by con- 
verting the sugar into starch. Accordingly, it is here that the formation of starch 
granules from the sugar is first noticed. When night falls or insolation ceases, the 
stomata close, the rate of evaporation slackens greatly, and then the concentrated sap, 
bathing the starch granules accumulated during the daytime, must become diluted by 
admixture with the dilute sap which continues to circulate owing to root pressure. 
This results in the solution becoming premaximal, so that the tendency towards increased 
conductivity determines the hydrolysis of the starch and its reconversion into 
suo-ar. 

To such speculation, unaccompanied by direct experimental evidence, a greater or 
less degree of plausibility can at best be conceded. More is not claimed for it at 
present. It is advanced here, in the hope of drawing the attention of botanists and 
physiological chemists to a line of investigation which, being based on a recognition of 
the importance of the actually occurring local variations in sap concentration, promises 
to be very fruitful. 

If we turn from the consideration of the metabolism in foliage leaves and other 
green parts of the higher plants to the processes belonging to the filling and ripening 
of seeds, we find analogous reversible reactions, but under conditions more amenable to 
direct experimental investigation, for here the reversals are not of diurnal, but are, as 
a rule, of annual occurrence. Anabolic polymerisations, condensations, and dehydra- 
tions characterise the filling and ripening of seeds, and these changes are certainly 
accompanied by the drying up and concentration of the sap. In presence of the 

* Brown and Morris, « The Chemistry and Physiology of Foliage Leave?," Chem. Soc. Jour., 1893, p. 637. 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 6). 21 



134 



PROFESSOR JOHN GIBSON ON 



necessary water, such reactions give place, during the germination of seeds, to 
catabolic hydrolysis, as in food digestion generally, both by plants and animals. 

In this direction the author has obtained what he regards as conclusive 
experimental evidence of the usefulness of the hypothesis of a tendency towards 
increased specific conductivity in relation to plant chemistry. This evidence bears 
upon the two classical examples of catabolic enzymatic reactions in seeds which 
are afforded by the splitting up of the glucosides, amygdaline and potassium 
myronate. By the use of the hypothesis the following simple experiments were 
suggested, and their results correctly predicted. The experiments consisted simply 
in mixing ordinary mustard powder and crushed bitter almonds respectively with 
the salt solutions which are given in Table III., some of which are premaximal and 
some ultramaximal, and then observing whether the characteristic pungency of allyl 
sulphocyanate or the smell of benzaldehyde became perceptible, and in what time. 

Table III. 



Relation to 
Maximal 

Solution. 



Ultra 
Pre 
Pre 
Pre 
Pre 
Pre 
Pre 
Pre 
Pre 
Pre 

Ultra 
Pre 

Ultra 
Pre 

Ultra 
Pre 



Electrolyte. 



Lithium chloride 

Sodium chloride (saturated) 

Potassium chloride (saturated) 

Potassium bromide (saturated) 

;) >> 

Potassium iodide (saturated) 

>> >) 

Potassium acetate . 

Magnesium chloride 

Calcium chloride 



Per cent. 



23-90 
1262 
26-40 
H-40 
24-86 
13-41 
40-28 
23-21 
58-24 
36-74 
46-74 
26-05 
23-81 
12-98 
28-84 
14-85 



r. 


10 3 K * 


10 3 K. 


10 v 


5-6 


168 


161 


7 


3-0 




140 




4-5 




216 




2-5 




161 




33 




330 




1-8 




180 




34 




392? 




1-9 




220? 




3-5 




460? 




2-2 


... 


260? 




5-4 


130 


90 


40 


3-0 




128 




5-0 


140 


134 


6 


2-7 




124 




5-2 


178 


171 


16 


2-7 




150 


... 



* Approximate values from graphical interpolation. 

The splitting up of amygdaline is certainly hydrolytic, and so probably is the 
splitting up of potassium myronate, although it is often represented by an equation 
which does not indicate hydrolysis. Certainly the presence of water is essential to 
both reactions. In each and every experiment the necessary water was present in large 
excess. 

The following question now arises : Which, if any, of these solutions should allow, 
and which of them should prevent or greatly retard, the interaction of the two gluco- 
sides with their corresponding enzymes, emulsin and myrosin ? 



MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 135 

Applying the hypothesis, the various solutions may be arranged into two groups : 
(A) Premaximal solutions and (B) Ultraraaximal solutions, the concentration of the 
solutions in group (A) being somewhat less, and that of those in group (B) somewhat 
greater, than the concentrations corresponding to maximum conductivity. 

It was predicted that the premaximal solutions would, permit the hydrolysis of the 
amygdaline and of the potassium myronate, but that the ultramaximal solutions would 
inhibit or greatly retard the hydrolysis and consequent splitting up of the glucosides. 
The results actually obtained justified the hypothesis in a very satisfactory manner. 
With crushed bitter almonds all the premaximal solutions gave the smell of 
benzaldehyde quite distinctly within five minutes or less after mixing ; and also gave, 
with ordinary English mustard powder, a distinct pungency.* Conversely, none of 
the ultramaximal solutions gave rise, on mixing, to the smell of benzaldehyde, or to 
any distinct pungency, on standing ten minutes. When the mixtures are allowed to 
stand in loosely covered vessels for twelve hours or so, the difference between the 
two sets becomes very marked. 

Without the hypothesis of a tendency towards increased conductivity as a directing 
factor it does not appear possible to predict the behaviour of the various systems. 
When the mixtures made with solutions belonging to group (B) — and in which the 
reactions were inhibited — were mixed with sufficient water to bring the solutions to a 
concentration distinctly lower than that corresponding to maximum conductivity the 
reactions were found to proceed readily, even after long standing. This demonstrates 
that the enzymes are not rendered permanently inactive, and completes the proof 
that the tendency towards increased conductivity is a real directing factor in these 
cases. 

Another interesting observation bearing on the behaviour of sulphuric acid may be 
cited here. 

Professor A. J. Brown has discovered that the covering of the seeds of Hordeum 
vulgare acts as a peculiar semi-permeable membrane. He finds, for instance, that it is 
possible to concentrate dilute sulphuric acid by soaking the barley grain in the dilute 
acid, for the membrane, when undamaged, permits the passage of water, but not of 
sulphuric acid, into the interior of the grain. 

The preceding examples suggest an answer to the following question : What is the 
maximum concentration of sulphuric acid which can possibly be reached in this way ? 
The answer is : Not beyond the concentration corresponding to maximum conductivity, 
i.e. 30 per cent. H 2 S0 4 , for, as has been seen, the sulphuric acid tends towards increased 
conductivity, so that when the concentration reaches that corresponding to maximum 
conductivity, the tendency will oppose further concentration, as that would imply a 
decrease in the specific conductivity of the solution, which would thereby become ultra- 

* This statement requires modification in one case only, viz. so far as mustard and saturated solution of 
potassium iodide are concerned, for this solution has a marked disintegrative effect and appears to produce a deeper- 
going change which masks the tendency of the premaximal system. 



136 MAXIMUM SPECIFIC ELECTRICAL CONDUCTIVITY IN CHEMISTRY. 

maximal. Professor Brown states further # that " it was found that sulphuric acid could 
not penetrate into the grain, not only from volume normal solutions, but also from 
solutions containing 9, 18, or even 36 grammes of acid per 100 c.c. In the case of the 
seeds immersed in the strongest acid, however, the interior remained dry, presumably 
because the power of the seed contents of imbibing water was insufficient to overcome 
the osmotic pressure of the liquid." 

"The vitality of the embryos was not destroyed by steeping the seeds in the acid 
solutions ; when placed under suitable conditions they all germinated." 

Now, 36 grams of sulphuric acid per 100 c.c. corresponds to a percentage of 297 
H 2 S0 4 , which is only 0'3 per cent, below that corresponding to maximum conductivity. 

It is suggested that in plant metabolism the tendency towards increased specific 
conductivity is a directing factor, both in the case of foliage leaves and of seeds. 
Increased specific conductivity results from the progress of the reactions in the one 
direction or in the other, according as the sap has the character of an ultramaximal, 
or of a premaximal solution. It is important in this connection to note that increase 
in specific conductivity does not solely result from the actual addition or removal of 
water" from the sap solutions, but that it also results from polymerisations or con- 
densations especially of the aldol type in submaximal, sub-premaximal, and sub- 
ultramaximal solutions, t 

It is hoped in a subsequent paper to develop the application of the hypothesis to 
chemical systems generally, and especially to those occurring in plant and animal 
metabolism. 

Several conjoint papers have been, or will shortly be, published bearing on different 
points related to the inquiry which is summarised in this paper. 

I take this opportunity of thanking those who in one way or another have given 
me kind assistance. My thanks are very specially due to Mr Andrew King, F.C.S., 
for long-continued and invaluable analytical assistance. I also gratefully acknowledge 
the receipt of a Research Grant from the Carnegie Trust. 

* Proc. Roy. Soc, 1909, vol. lxxxi. p. 82. 
+ See page 124 et se<j. 

Hekiot-Watt College, 
Edinburgh. 



( 137 ) 



VII. — Nuclear Osmosis as a Factor in Mitosis. By A. Anstruther Lawson, Ph.D., 
D.Sc, F.L.S., F.R.S.E., Lecturer in Botany, University of Glasgow. (With 
Four Plates.) 

(Read June 19, 1911. MS. received July 25, 1911. Issued separately November 7, 1911.) 

The recent important work of Farmer and Digby (1910) on the cytology of certain 
varietal and hybrid ferns has raised again the interesting and absorbing question of 
the mechanism of mitosis. This paper constitutes a valuable addition to the literature* 
which has bulked so largely during the last few years, and which has been instrumental 
in establishing many interesting and important facts regarding the greatest of all 
cytological problems — namely, nuclear division. The interest of this later contribution 
lies, not so much in the actual observations which these writers have jointly recorded 
— although the facts revealed are important in that they confirm the work of other 
writers and extend the range of observation, — but rather in the theory which they 
have deduced from their results and which they have put forward to account for the 
factors at work in the living cell, — factors which they believe to be concerned in the 
formation of the achromatic spindle. 

The observations made in this more recent paper confirm the generally accepted 
view that the achromatic spindle is formed from a differentiation of the cytoplasm — 
a differentiation which, on account of the active role it is believed to play in mitosis, 
has been called Mnojrfasm by Strasburger. They also confirm the view that — with 
the exception of some unessential variations in the distribution and orientation of the 
kinoplasm in the prophase — the process of spindle development is fairly uniform 
throughout the vascular plants. The establishment of such uniformity is valuable, 
because it follows that the factors concerned in the process would also be uniform, and 
any theory accounting for such factors would necessarily be far-reaching in its application. 

From the facts and views expressed in the above literature we may briefly 
summarise the mitotic process as follows : — It seems that with the change of the 
spireme into definite bivalent chromosomes, the cytoplasm in the immediate vicinity 
of the nuclear membrane becomes differentiated into a series of delicate fibrils (kino- 
plasm), forming a weft which more or less completely surrounds the nucleus. In 
somatic cells this weft becomes raised up from the nuclear membrane at opposite 
points in such a fashion as to form conical-shaped caps of kinoplasm, which, on account 
of their position, are known as polar caps. The fibrils composing the caps converge at 

* Strasburger (1882, 1895, 1905, 1907); Belajefj? (1894); Farmer (1893, 1895, 1905, 1910); Osterhout 
(1897) ; Juel(1897) ; Sargant (1897); Lawson (1898, 1900, 1903); Debski (1897); Williams (1899); Byxbee(1900); 
Smith (1900) ; Nemec (1899) ; Mother (1897, 1898, 1907) ; Miyake (1905) ; Allen (1903, 1905) ; Overton (1905, 
1909) ; Berghs (1904, 1905) ; Gregoire (1903) ; Davis (1899, 1909). 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 7). 22 



138 MR A. ANSTRUTHER LAWSON ON 

points which are directly opposite one another, so that the spindle is bipolar from its 
inception. This is also true in some cases for the heterotype spindle, but as a rule in 
the latter, the weft of kinoplasm pushes out at a number of places and the fibrils converge 
at several points which are not necessarily opposite one another. The result is that a 
number of fibril sheaves are developed which are of conical shape, and the spindle at 
this prophase may be tripolar, quadripolar, or multipolar. By the coalescence of these 
cones, however, the spindle eventually becomes bipolar and has the same symmetrical 
form as the somatic spindle. With the organisation of the cones in the prophase — 
whether the spindle is bipolar or multipolar — it is believed that the nuclear membrane 
breaks down and disappears. The fibrils of the spindle now push into the nuclear area, 
and with their free ends become attached to the chromosomes. The latter become 
arranged at the equator, forming the characteristic equatorial plate, and we now have 
what is known as the metaphase. Each bivalent chromosome separates into two 
daughter chromosomes, each of which moves to opposite poles of the spindle. It is 
believed by some cytologists that it is the contraction of the fibrils attached to the 
chromosomes which accomplishes not only the separation of these bodies, but also the 
migration of the two halves to the poles. 

These, in brief, are the conclusions reached by many investigators whose observations 
have extended over a wide range of types of vascular plants. They are the views that are 
commonly held at the present time, and they have in the main been sustained by the 
recent work of Farmer and Digby (1910). These writers, however, have done more. 
They have drawn some theoretical conclusions from their observations in an attempt to 
account for the factors concerned in the various changes and movements in the cell 
expressed in the achromatic figure. Following the work of Professor Marcus Hartog, 
they have come to the conclusion that an explanation of these changes and movements 
may be found in the electrical conditions of the cell. They point out the remarkable 
manner in which the sheaves of fibrils, during the prophase, diverge in the proximity 
of the chromatin-charged linin, and that these are so repelled by each other that they 
press out equidistantly at the periphery of the cytoplasm. This condition appeals to 
them as convincing evidence in support of their hypothesis that the linin, with its 
contained chromatin, by virtue of the chemical changes involved in its metabolism, has 
brought about an electrical condition of opposite sign, similar in each of the spindle 
cones. This hypothesis seems to them to be in harmony with the fact that the dis- 
appearance of the nuclear membrane is closely associated with the spreading of the 
chromosomes beneath it just before their retrogressive movement to the equator, whilst 
the spindle poles have shifted away from the nuclear surface. For these and other 
reasons Professor Farmer lends his support to the view that electrical conditions in 
the cell are not only responsible for the form of the spindle, but for its very existence 
— a view frequently put forward, but not generally accepted. 

Without attempting an analysis of this hypothesis, T should like to point out some 
difficulties which 1 have experienced in interpreting the observations which have been 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 139 

recorded in the process of spindle development as ordinarily accepted, and which I 
have briefly outlined above. 

In the first place, the cytoplasm — presumably in the form of a reticulum — in the 
immediate vicinity of the nuclear membrane becomes transformed into a series of 
delicate threads or fibrils, which forms a dense weft about the nucleus. There seems 
not much doubt that these threads are really transformed cytoplasm, and this is the 
generally accepted view. It is believed that these kinoplasmic threads eventually 
become the fibrils of the mature spindle. But just how that transformation has been 
brought about and the factors responsible for the changes are questions which have 
never been satisfactorily answered. The kinoplasmic idea of Strasburger offers no 
explanation — it is simply descriptive. The hypothesis of electrical conditions and 
phenomena of induction as being factors in these changes is vague and difficult to 
comprehend. This idea, which has again been advocated by Hartog (1910), is no 
doubt very suggestive, but unfortunately it is based more upon appearances than upon 
knowledge of such electrical conditions in the living cell or upon the knowledge of the 
effect of such conditions upon living protoplasm. 

Another difficulty — and in my opinion the greatest difficulty in the whole process of 
mitosis— that requires some explanation is the breaking down, collapse, and disappear- 
ance of the nuclear membrane. Many investigators (including the present writer) have 
either described or figured the breaking down of the nuclear membrane at a time when 
the multipolar stage has been reached, or in somatic cells when the polar caps have been 
completely formed. This stage has been so frequently described that its actual occurrence 
has never been questioned. Now it is a fact of general acceptance that the nuclear 
membrane is cytoplasmic in nature (Peirce, 1903) — it is, in fact, the inner limiting 
layer of cytoplasm, and it is a membrane in virtue of its having come in contact with 
the karyolymph (Pfeffer, 1890; Gregoire, 1903; Lawson, 1903; Yamanotjchi, 
1906). It is not only a plasmatic membrane, it is permeable and consequently osmotic. 
The nucleus itself is an osmotic system, and its membrane is the main essential element 
in that system. It should be remembered that at the time the membrane is supposed 
to break down, the nuclear cavity, while considerably under its maximum size, is 
nevertheless very large, and the membrane is under considerable tension, due to the 
high osmotic pressure of the karyolymph. Furthermore, it is well known (Lawson, 
1903; Gates, 1908; Farmer and Digby, 1910) that the membrane is capable of 
increasing or diminishing its surface area. This is clearly demonstrated in any spore 
mother-cell where changes in the size of the nuclear cavity may be easily observed. So 
that, in this sense, the membrane is both stretchable and elastic. The point that I am 
endeavouring to make is, that it is difficult to understand why a membrane with these 
properties should break down under the conditions of spindle formation. During the 
prophase the karyolymph is undoubtedly exerting a considerable pressure upon the 
membrane. This is clearly indicated by the almost spherical form and turgid appear- 
ance of the nucleus at this time. One would expect a less violent process than breaking 



140 MR A. ANSTRUTHER LAWSON ON 

down and collapse under such circumstances. Nevertheless, the actual breaking down 
has been described and figured by several close and accurate observers. Without 
attempting to question the accuracy of these records, I should like to point out a fact 
which will become more obvious later in this parjer. It is, namely, that at the time 
when the membrane is reported to break down, the nuclear cavity has diminished con- 
siderably in size— that there is less karyolymph than formerly. A rational explanation 
for this change is a variation in the osmotic relations — that a part of the karyolymph 
has diffused into the cytoplasm by exosmosis. Under such circumstances it is not 
difficult to understand the collapse of the nuclear membrane upon the sudden 
application of fixing reagents. After a careful re-examination of my own preparations, 
I have not only been convinced that this is really the cause of the breaking down of 
the nuclear membrane, but also that in the normal living cell the membrane does not 
break down. What would happen if such a breaking down were possible in the living 
cell ? Clearly the surrounding cytoplasm would be at once exposed to the large 
remaining body of karyolymph, which is a watery fluid. By the very reason of that 
exposure and contact one would expect a new membrane to be precipitated immediately. 
The new membrane would be precipitated so simultaneously with the break that the 
latter could not be detected. For does not the nuclear membrane exist as a membrane 
by virtue of its contact with the karyolymph ? (Pfeffer, 1890 ; Lawson, 1903 ; Gates, 
1907, Yamanouchi, 1906). The killing of the cytoplasm by the application of fixing 
reagents would prevent the precipitation of a new membrane, so that any rupture caused 
by such reagents would show in fixed material. 

The growing or pushing in of the fibrils from the base of the cones into the nuclear 
area when the nuclear membrane is supposed to have vanished has also offered some 
difficulties. It is generally admitted that these fibrils are nothing but modified 
cytoplasm, and as such they are viscous and semi-fluid. It seems improbable that such 
fine, delicate cytoplasmic threads should traverse the clear remaining body of nuclear 
fluid — threads which at one moment are reported to be elongating towards the periphery 
of the cell, and in the next, upon the disappearance of the nuclear membrane, are observed 
to traverse or elongate in the opposite direction. Surely something more than the 
vague mystery of " electrical conditions " is needed to account for such extraordinary 
changes. But, as Professor Farmer admits, " the time has not yet arrived when it will 
be possible to give an explanation of these cellular changes that will prove satisfactory 
from a physical point of view." 

These threads seem not only endowed with the power of traversing the clear watery 
fluid of the nuclear area, but it is reported that they become attached to the chromo- 
somes by their free ends. Now it is a fact that all of the chromosomes become attached 
to fibrils — none of them escape. It seems also that the distribution of the fibrils is 
fairly uniform among the chromosomes. That is to say, the number of fibrils that 
become attached to the various chromosomes is approximately the same, there being 
no striking difference in the size of the fibril sheaves attach to each chromosome as 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 141 

we see them in the mature spindle. Moreover, these fibrils attach themselves to 
opposite sides of each bivalent chromosome, and the sheaf on one side is approximately 
equal to that on the other. The difficulties here presented are quite obvious. In the 
first place, an explanation is needed to account for the actual attachment of the free 
ends of these fine, delicate, in-growing fibrils to the chromosomes, and if they do so 
attach themselves, this implies a selective power either on the part of the fibrils or of 
the chromosomes both in regard to numbers and position of attachment. 

Another series of difficult interpretations is to be found in reference to the 
coalescence of the cones of the multipolar stage which results in the bipolar figure. It 
will be remembered that the apices of the cones may be some distance apart, but it is 
believed that they eventually approach one another at opposite points. Such approach- 
ment and coalescence implies a movement. Such a movement would necessarily be through 
the undifferentiated reticulum of the cytoplasm. It should be remembered that certain 
of the apices of the cones are so far apart that they pass through an angle of 45° in the 
accomplishment of this fusion. In this connection Professor Farmer calls attention to 
the aggregation of the nuclein-charged linin beneath the points of origin of the cones, 
and he sees in this an indication that there exists a causal connection between the two 
phenomena ; and just as the chromatin at this early stage appears to determine the 
formation of the four or more poles, so now the chromosomes again appear to be the 
active agents in effecting the resolution of the quadripolar into a bipolar arrange- 
ment. But just how this is brought about Professor Farmer does not explain. The 
fact is, the idea of approachment and eventual coalescence of the cones has been 
based upon appearances only. A careful search of the literature and of my own 
preparations has failed to reveal the slightest indication that the cytoplasmic reticulum 
has been in any way disturbed by such a coalescence. No real evidence has been 
recorded to show that an approachment of the cones really occurs. I shall attempt to 
prove in the following pages that the idea of a lateral movement and coalescence of 
the cones is a misconception. 

Perhaps the most striking feature of the whole mitotic process is the constancy and 
uniformity of the equatorial plate in the metaphase. Here the chromosomes always 
arrange themselves in the same plane at the equator of the spindle. This plane 
constitutes a dividing line between two series of fibril-sheaves which extend in 
opposite directions and generally converge at the poles. It is this constant position 
of the chromosomes which gives the spindle its character. Up to the present no 
adequate explanation has been offered to account for the factors concerned in the 
arrangement of the chromosomes in this peculiar fashion which results in so striking 
a feature of mitosis. To say that the equatorial plate is due to electrical conditions 
only confuses the issue, at least until more is known of such electrical conditions 
and their influence upon protoplasmic bodies. It is just as profitable to say that 
it is due to cell-polarity, the determining factors of which are unknown so far as the 
vascular plants are concerned. 



142 MR A. ANSTRTJTHER LAWSON ON 

It is believed by some that the meaning and function of the achromatic figure is to 
accomplish the separation of the chromatin substance of the mother nucleus into two 
equal portions. It is believed that by the contraction of the fibrils which undoubtedly 
become attached, the daughter chromosomes are not only separated from one another, 
but are actually pulled to opposite poles of the spindle. Here again I find some 
difficulties. In the first place, the poles of the bipolar spindle rarely extend as far as 
the cell-wall, and consequently if such actual pulling of the chromosomes takes place, 
there appears to be no stationary resistance to pull against. And then again, from an 
examination of many preparations I find that the daughter chromosomes invariably 
pass beyond the region of the pole of the spindle. These and other facts do not lend 
support to the idea that this movement is accomplished by the contractility of the 
connected fibrils. 

And finally the question presents itself: If these fibrils of the achromatic spindle 
have no concern in the separation of the daughter chromosomes, what is the reason for 
the existence of such a complicated mechanism ? 

In view of the many difficulties and inconsistencies mentioned above, is it not 
possible that the development and function of the achromatic figure have been mis- 
interpreted ? Is it not possible that the so-called "mechanism of mitosis" represents 
the passive results and effects of movements rather than an active agent in such 
movements ? After a careful, comprehensive study of a wide range of types, I am 
forced to answer these questions in the affirmative. I have also been convinced that 
the cause of so many difficulties lies in the fact that a series of important and critical 
stages in spindle formation has been overlooked — stages which throw an entirely new 
light on the problem, and which will necessitate a revision of the accepted views and 
interpretations of nuclear phenomena. 

These stages are to be found in the later prophase, preceding the organisation of 
the equatorial plate. They are stages concerning the fate of the nuclear membrane. 
Contrary to my earlier observations, as well as to the observations of the majority of 
cytologists, I find that the nuclear membrane does not break down, but, on the contrary, 
behaves as a permeable plasmatic membrane should behave under varying osmotic 
relations. The discovery of these stages not only clears up a doubtful point which I 
mentioned above in regard to the prophase, but is far-reaching in its importance. It 
goes to prove that osmotic conditions rather than electrical conditions are active factors 
concerned in the formation of the achromatic spindle. This I shall attempt to 
demonstrate from a study of the spore mother-cell of Disporum, Gladiolus, Yucca, and 
Hedera, also the vegetative cells in the root tip of Allium. 

Disporum. 

In fig. 1 we have represented a median section of a microspore mother-cell of 
Disporum Hookeri. It is taken at a time when the nuclear cavity is just about its 
maximum size, soon after the growth period, which up till recently was known as 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 143 

"synapsis."* Within the nucleus there are five bivalent chromosomes almost com- 
pletely organised, and a large spherical nucleolus. Surrounding the enlarged nuclear 
cavity there is a very distinct membrane, which shows every indication of being under 
pressure of the karyolymph. The surrounding cytoplasm, which appears as a relatively 
narrow zone in section, is densely but finely granular and of a reticulate structure. 
Enveloping the whole is a fairly thick and apparently firm cell-wall. It will be seen 
that the nucleus occupies more than half the cubical volume of the cell. At this time 
the cytoplasm is perfectly uniform, with no indication of kinoplasm t having been 
differentiated. 

In fig. 2 we have represented a little later stage. Here it will at once be noticed 
that the nuclear volume has diminished considerably in size, that the amount of 
karyolymph present is not so great as that represented in the first figure. It will also 
be noticed that the width of the cytoplasmic zone has correspondingly increased. 
There is also to be seen a distinct modification in the structure of the cytoplasm, 
especially in the immediate vicinity of the nucleus. The cytoplasm in this region has 
taken on the form of fine, delicate threads which appear to radiate out a short distance 
from the nuclear membrane towards the periphery of the cell. In fig. 3 we represent 
a cell where the diminution of the karyolymph has continued to such an extent that its 
cubical volume is less than half that shown in fig. 1. Here also we see the cytoplasm 
in the form of radiating threads ; but these latter are much more sharply defined and 
longer than those shown in fig. 2. This is really the stage where the nuclear membrane 
is reported to break down in other types. But, as shown in figs. 4 and 5, this certainly 
does not occur in Disporum. It should be noted in passing that in these cells 
the characteristic multipolar arrangement is not so sharply defined as in many other 
plants, the kinoplasm taking on the form of radiations which form small tuft-like 
sheaves. The nuclear membrane remains intact, and, as indicated in figs. 4, 5, and 6, 
it gradually closes in the nuclear cavity with the diminution of the karyolymph. 
This latter, although now very much reduced in quantity, apparently still exerts a 
considerable pressure on the enveloping membrane, for the nuclear cavity is still 
spherical and turgid. 

These changes in the dimensions of the nuclear cavity are evidently gradual, but 
the point of interest is that they are always accompanied by a corresponding increase 
in the differentiation of the surrounding cytoplasm ; for it is evident from figs. 3, 4, 5, 
and 6 that, as the nucleus becomes smaller and smaller, the cytoplasmic radiations 
become longer and more sharply defined. In fig. 7 we have a most interesting stage. 
Here the karyolymph has diminished to such an extent that the chromosomes appear 
crowded about the nucleolus and the clear nuclear sap is hardly visible. The nuclear 
membrane is in close touch with the chromosomes for the greater part of its surface. 

* Lawson, A. A. (1911). 

t I intend using the word " kinoplasm ; ' throughout this paper as a convenient term, but not, however in the 
sense in which it was first applied by Strasburger. 



144 MR A. ANSTRUTHER LAWSON ON 

We have here, from figs. 3 to 7, a series of stages which, as far as I know, has never 
before been described. That it represents early phases in the development of the 
spindle is, I think, quite obvious. It also demonstrates that with the gradual decrease 
in the nuclear volume there is a corresponding increase in the cytoplasmic volume. It 
should be remembered that throughout these stages, as well as those immediately pre- 
ceding, the cell is surrounded by a firm cell-wall of considerable thickness, to which the 
peripheral cytoplasm is closely associated, and that a perceptible diminution in the 
volume of the ceZZ-cavity is inconceivable. It should also be remembered that the 
nuclear membrane is part of the cytoplasm. With these facts in mind it is quite clear 
that in the stage represented by fig. 7 the cytoplasm fills a cubical space which is more 
than double that shown in fig. 1 , but there is no change in the dimensions of the cell 
as a whole. That varying osmotic relations constitute a causal factor in these trans- 
formations seems to me a safe and rational assumption. Everything necessary to 
promote osmotic diffusion is present. There is a permeable membrane and substances 
of different chemical composition and presumably of different density on either side of 
it. It is therefore not difficult to understand the gradual diminution of the karyo- 
lymph, as shown in these figures, on the basis of osmotic diffusion. The karyolymph 
has passed out into the cytoplasm by exosmosis. 

All of these circumstances bring about a condition where a limited amount of 
cytoplasm of reticulate structure is obliged to occupy a cubical space which is gradually 
increased by reason of the diminution of the large nuclear space. This necessarily sets 
up a state of tension in the cytoplasm — a tension sufficient to cause a readjustment and 
changed configuration in the reticulate form of the cytoplasm ; a change to the form 
of threads or fibrils which are drawn out from the reticulum by the receding nuclear 
membrane. With the closing in of the nuclear membrane about the chromosomes the 
cytoplasm must follow the membrane, and the changed configuration of the cytoplasm 
would first make itself evident in the region of the membrane. On examination of the 
serial stages represented in figs. 1 to 7, the conclusion is irresistible that this is really 
what has happened. The kinoplasmic fibrils are drawn-out threads of cytoplasm — 
drawn out by reason of the inward movement of the membrane. 

But now, what follows the stage represented in fig. 7 ? In fig. 8 we have a later 
condition, where the karyolymph and membrane are no longer visible. This invisi- 
bility, however, does not prove that these elements of the nucleus no longer exist. They 
are both present — the one very much reduced and saturating the chromosomes, and the 
other completely enveloping each chromosome. So that we now have, not a single 
osmotic system as formerly, but as many osmotic systems as there are bivalent 
chromosomes ; and in this case there are five. 

This closing in of the membrane about each bivalent chromosome is doubly 
interesting. It not only establishes a number of osmotic systems which are more or 
less independent of one another, but it clears up a doubtful matter, mentioned above, 
as to the manner in which the kinoplasmic threads become attached to the chromosomes. 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 145 

These threads from their very first appearance are moored to the membrane — being 
a continuation of the same cytoplasmic substance — and that association continues 
throughout all of these stages. As the membrane recedes, the threads are drawn in 
with it, and so it comes about that as the membrane becomes closely applied to the 
surface of the chromosomes, each of these bodies is furnished with its own system of 
fibrils, which to all intents and purposes are attached, as shown in fig. 9. This seems to 
me a much more rational method of attachment than the commonly accepted view that 
the threads push into the nuclear area and attach themselves individually to the 
chromosomes with their free ends. It also accounts for the more or less equal 
distribution of the fibrils among the chromosomes ; for these threads would continue 
their fairly uniform occurrence over the osmotic surfaces, that is, over the membranes 
enveloping each chromosome, and as these latter are approximately the same size and 
shape, the amount of fibrils furnished to each chromosome would be about equal. 

It is obvious that as osomotic diffusion progresses and the nuclear vacuole becomes 
smaller and smaller, there will follow not only a corresponding shifting of the lines of 
tension as expressed in the kinoplasmic threads, but also an acceleration of the changes 
occasioned by such a shifting. So that the interval between the stages represented in 
figs. 6 and 9 would be a very brief one, and such critical periods would not be very 
frequently observed. It is also obvious that, as the karyolymph becomes gradually 
exhausted by the continued osmotic diffusion, it eventually becomes no longer visible 
as a clear nuclear sap. This transition is illustrated in figs. 7 and 8. As the condition 
shown in fig. 8 is approached there naturally follows a further readjustment of the lines 
of tension. In the earlier stages, as shown in figs. 3, 4, 5, and 6, these lines of tension 
radiate out from the nuclear membrane with the nucleus itself as a centre ; but when we 
reach the stage represented in figs. 8 and 9, this condition no longer prevails. The lines 
of tension have readjusted themselves to meet the new condition. Each bivalent 
chromosome becomes the centre of a system of fibrils, but on account of the crowded 
position of the chromosomes (fig. 8) at this time, the radiations are not so regularly 
disposed as in the earlier stages. Now, I do not for a moment believe that these 
radiating threads — which merely express lines of tension — move individually or 
collectively through the cytoplasm. The apparent change in their position is due to 
the relaxing of the tensions along certain lines and establishment of new tensions 
along others. In other words, while the lines of tension do shift, the actual 
threads of cytoplasm do not. These latter withdraw or reappear, according to the 
shifting of the position of the osmotic surfaces, that is, the membrane enveloping 
each chromosome. 

Another point of great interest in this connection, and revealed in the stage shown 
in fig. 8, is the distribution of the fibrils over the surface of the chromosomes. Here it 
will be seen that the chromosomes are longer than broad, and the line of division 
between the two daughter chromosomes is parallel with the long axis. If the chromo- 
somes were perfectly spherical, one would expect a uniform and symmetrical radiation 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 7). 23 



146 MR A. ANSTRUTHER LAWSON ON 

of the fibrils, but, as seen in fig. 8, the lateral surface of each chromosome is furnished 
with many more fibrils than the end surface, and for this reason the disposition of the 
fibrils is not symmetrical. Also, the close juxtaposition of the chromosomes would aid 
in the prevention of symmetry. This unsymmetrical arrangement, however, is of very 
short duration. The greater surface from which project the larger number of fibrils 
apparently dominates the direction of the lines of tension, and in this readjustment the 
bivalent chromosomes become less crowded and each one becomes suspended between 
two sheaves of fibrils, as shown in fig. 9. 

The state of tension set up in the cytoplasm by the gradual diminution and final 
vanishing of the nuclear vacuole now finds an expression in two conical-shaped sheaves 
of kinoplasmic threads which appear on opposite sides of each bivalent chromosome, 
so that, as shown in fig. 9, each of these bodies is provided with an independent 
miniature spindle, and these lie parallel to one another. Now, as the surfaces which 
form the base of attachment for the fibrils are equal, so the two sheaves on the opposite 
sides of each chromosome are also equal, and the chromosome is thus suspended at the 
equator. Applying this to all five chromosomes, and taking into consideration the 
more or less parallel arrangement of the main lines of tension, we have here an exceed- 
ingly suggestive explanation for the organisation of the equatorial plate. But in this 
connection it should be remembered that we know little or nothing in regard to the 
problem of cell polarity in the vascular plants, and the plane occupied by the chromo- 
somes during the metaphase is too closely involved in this problem to justify my 
offering the above as an adequate explanation. I have no hesitation, however, in 
expressing my belief that the osmotic surfaces enveloping each chromosome are 
determining factors in the suspension of these bodies between two sets of fibril sheaves 
of fairly equal size. 

From a study of dividing cells in Disporum and other types which I will mention 
later, no evidence was found to support the view, held by Strasburger and others, 
that the daughter chromosomes are drawn to the poles by the contraction of the spindle 
fibrils. That these fibrils shorten and thicken with the movement of the chromosomes 
towards the poles is no doubt quite true, but this is not sufficient proof that they are 
actually engaged in the pulling of these bodies to opposite ends of the spindle. I believe 
that this shortening and thickening is due to the relaxing of the tension in these 
fibrils as the chromosomes move to the poles, and the fibrils merely act as guide lines 
with no pulling force. Although the tension in the fibrils between the daughter 
chromosomes and the poles of the spindle is thus relaxed, as shown in fig. 10, the 
tension in the cytoplasm as a whole has not been relaxed, for we see new lines of 
tension expressed in the fibrils stretching between the pairs of daughter chromosomes. 
This is quite clear in figs. 10 and 11. There has merely been a shifting of the lines of 
tension caused by the movement of the chromosomes to the poles. It should be noted 
in passing that the daughter chromosomes move beyond the actual poles of the spindle. 
This feature was observed not only in Disporum, but in many other types, and it 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 147 

supports the statement made above that the contractility of the fibrils is not the 
cause of the separation of the daughter chromosomes. 

In the stage represented in fig. 1 1 we see the two groups of daughter chromosomes 
lying at opposite ends of the cell, with a series of cytoplasmic threads stretching 
between them. My interpretation of these fibrils at the close of the anaphase is that 
they represent the same state of tension by the spindle in the metaphase and the 
radiating kinoplasm of the prophase — a tension caused in the first place by the 
diminution and final vanishing of the nuclear vacuole. 

The daughter chromosomes forming the two groups now become more or less 
intimately united — a condition in which it is difficult to distinguish the individuals. 
With this massing together each chromosome loses its compact and homogeneous 
appearance by becoming vacuolated. This vacuolation of the daughter chromosomes 
was first described by the writer in the case of Passifiora and Equisetum (Lawson, 
1903), and was later confirmed by other writers (Gregoire et Wygaerts, 1903; 
Yamanouchi, 1906 and 1910; Gates,' 1907). It would seem that this is a constant and 
normal occurrence in the organisation of daughter nuclei. Yamanouchi (1910) regards 
this vacuolisation as either a secretion from the chromosomes or a dissolution of portions 
of them into liquid. In the light of the present researches I am inclined to regard the 
presence of these lacunae or vacuoles within the chromatin as due to the diffusion of 
endosmosis. In this connection I see no reason why the chromosomes should not still 
possess the plasmatic membrane with which they were enveloped in the later prophase 
(see p. 144 above). This vacuolisation proceeds until the membrane extends beyond 
the limits of the chromatin, as shown in fig. 12, and eventually each daughter nucleus 
consists mainly of a single large vacuole filled with karyolymph. As these two nuclear 
vacuoles increase in size there naturally follows a general relaxation of the tension set 
up in the cytoplasm at the prophase, and this relaxation is expressed in the loose curved 
appearance of the threads stretching between the daughter nuclei, as shown in fig. 13. 

Gladiolus. 

In contrast to the conditions found in Disporum, the spore mother-cells in Gladiolus 
are very large ; and they have the additional interest of passing through a series of 
typical multipolar stages in the course of spindle formation, which has been described 
as occurring so frequently in the vascular plants. 

In fig. 1 4 we have represented a median section of a miscrospore mother-cell of the 
common garden species of Gladiolus. The large spherical nucleus occupies about half 
the volume of the cell-cavity. It is apparently at or near its maximum size, and, judging 
from its turgid appearance, it is evidently under high osmotic pressure. The chromo- 
somes appear as curved or bent rods many times longer than broad. A large globular 
nucleolus is also suspended in the karyolymph. The nuclear membrane stands out in 
sharp contrast as the inner limiting layer of cytoplasm against the clear nuclear sap. 



148 MR A. ANSTRUTHER LAWSON ON 

Within the meshes of the cytoplasm there are numerous oil globules and other food 
granules. In other respects the cytoplasmic reticulum is quite uniform. The peri- 
pheral cytoplasm is in close association with a firm cell- wall of considerable thickness. 

In fig. 15 we represent a condition which shows unmistakable evidence of a 
diminution in the volume of the nuclear vacuole. Accompanying this diminution we 
find the first indication of the transformation of the cytoplasmic reticulum into kino- 
plasmic threads. Here it will be seen that the cytoplasm in the region of the nuclear 
membrane has undergone a modification which takes the form of a narrow weft of 
delicate threads. This weft does not appear to be uniform in its distribution about 
the nucleus, and differs from the early kinoplasmic zone described above for Disporum 
in that the threads do not form a system of radiations. It appears that as the nuclear 
membrane recedes with the diminishing of the karyolymph, the drawing out of the 
threads from the cytoplasmic reticulum is more marked at certain places than at others 
(figs. 15 and 16). At such places there are conical-shaped sheaves of threads produced, 
which impart an irregular and unsymmetrical form to the kinoplasm. The beginning 
of these sheaves is shown in fig. 15, and later stages are shown in figs. 16 and 17, 
where we have a typical multipolar arrangement. In the stages shown in figs. 16, 17, 
and 18, we find in each cell several large conical-shaped sheaves of fibrils whose wide 
bases are evidently continuous with the nuclear membrane, and whose outer or distal 
extremities taper out into more or less sharp points. The drawings are, of course, 
made from sections, and consequently the figures do not represent all of the cones 
developed in each cell. In fig. 16 there are three of these cone-shaped sheaves to 
be seen; in fig. 17 there are four; in fig. 18 three; and in fig. 19 two. A study and 
compilation of serial sections convinces me that the number of sheaves developed in 
these early stages is not an essential feature. In the matter of numbers they seemed 
to vary considerably in the different cells I have examined. The form and position of 
the kinoplasmic sheaves seem likewise a matter of no great importance, for, as I shall 
point out later, they are constantly changing during the prophase. The feature of these 
early stages that is of great importance, and one that should be noted as having an 
essential bearing on all the changes of the prophase, is that as these kinoplasmic 
sheaves develop and increase in size there is a corresponding gradual decrease in the 
volume of the nuclear vacuole. It will be seen that as the nuclear vacuole becomes 
smaller and smaller, the kinoplasmic threads become longer and more sharply defined. 
That there exists a causal relation between these two sets of changes I have no doubt 
whatever. If measurements be taken of the volume of the nucleus as indicated in the 
sections shown in figs. 20 and 21, it will be found that there has taken place a great 
reduction in the karyolymph— that its volume is now only about one-eighth of that 
shown in nucleus in fig. 14. As I pointed out above in the case of Disporum, we have 
here a limited amount of cytoplasm now occupying a cubical space which has been 
very much increased by reason of the diffusion of the karyolymph from the nuclear 
vacuole. The state of tension that necessarily follows finds an expression in the 



NUCLEAE OSMOSIS AS A FACTOR IN MITOSIS. 149 

drawing out from the cytoplasmic reticulum a series of fine, delicate threads. If we 
now examine closely the kinoplasm which has been thus differentiated, we will find that 
the changed configuration of the cytoplasm is much more sharply defined near the 
nuclear membrane, and as we follow the threads outward to the periphery, they 
gradually fade out and become lost in the reticulum. From this it would seem that 
the tension expressed in these fibrils decreases in proportion with the distance from the 
nuclear membrane. This is obviously beyond actual proof, but if it is true, as appear- 
ances seem to indicate, it may account for the conical shape of the groups or sheaves 
of fibrils, for these clearly show a like tendency to attenuate as they approach the 
periphery. 

A comparison of the serial stages shown in figs. 14 to 21 is sufficiently convincing 
that while there is a marked but gradual decrease in the volume of karyolymph there is 
no evidence whatever that the nuclear membrane breaks down, collapses, or disappears. 
On the contrary, the membrane remains intact throughout all of these stages. As 
indicated in figs. 19, 20, 21, and 22, the nuclear vacuole may not always retain its 
spherical form, and the shape of the membrane may vary accordingly, but even in such 
extreme cases as shown in fig. 22 the contour of the membrane in section may easily 
be followed. It has simply receded with the gradual diffusion of the nuclear sap. 

Now, the nuclear membrane not only remains intact throughout the prophase, but 
it continues to form the base of the lines of tension expressed in the drawn-out 
threads, for it is in reality a continuation of the same cytoplasmic substance as the 
kinoplasm. It would therefore naturally follow that, as the membrane receded with 
the diminution of the nuclear vacuole, the lines of tension would shift accordingly. 
Such a shifting does not mean the changing of the threads bodily from one position to 
another. It means the relaxing of the tension along certain threads, which would 
consequently fall back into the form of the original reticulum, and the setting up of 
new lines of tension, with the drawing out of new threads from the hitherto undiffer- 
entiated reticulum. In this fashion not only individual threads, but entire sheaves or 
cones may appear to assume different positions. I believe the conditions shown in 
figs. 20 and 21 represent transition stages in the shifting of the lines of tension, and 
consequently the apparent shifting in the position of certain cones or sheaves in this 
way. In both these figures there are portions of the cytoplasm which can be in- 
terpreted in no other way than transitions between a reticulum and kinoplasmic threads. 
I regard such demonstrations as very important, because I believe they suggest a 
rational explanation not only for the apparent changes in the position of the fibril 
sheaves, but also for the ultimate resolution of the multipolar into a bipolar arrange- 
ment. In figs. 17, 18, 23, and 24, it will be seen that the apices of the cones may 
be a considerable distance apart. It is generally believed that the apices approach 
one another and the cones ultimately coalesce in two groups. Such a movement of 
the cones bodily seems to me not only improbable but impossible, without some dis- 
turbance of the cytoplasmic reticulum. In the series of stages shown in figs. 16 to 



150 MR A. ANSTRTJTHER LAWSON ON 

25 there is no trace whatever of such a disturbance in the cytoplasm. There is no 
evidence of any sort to support the view that a movement and coalescence of the 
cones really occurs, except the fact that we eventually have a bipolar figure in place of 
one which was previously multipolar. Throughout all of these stages it will be seen 
that the nuclear membrane, which in reality constitutes the bases of the cones, is con- 
stantly moving inward as it closes in about the chromosomes, and as the nuclear 
vacuole becomes smaller and smaller the area of the membrane becomes less and less. 
With these changes there necessarily follows a corresponding shifting in the lines of 
tension ; so that during the entire prophase there is a constant changing in the dis- 
tribution of the kinoplasm. The ultimate bipolarity of the spindle is therefore not 
brought about by the approachment and coalescence of the cones in two groups, but 
by the shifting of the osmotic surfaces which form the bases of the lines of tension 
represented in the threads of kinoplasm. 

In fig. 22 we have a condition where the karyolymph has been reduced by diffusion 
to such an extent that the chromosomes have become crowded together around the 
nucleolus. The nuclear membrane is now in close touch with several of the chromo- 
somes. In the lower part of the figure one of the chromosomes seems to be already 
partly enfolded by the membrane. That this enveloping process continues until the 
membrane becomes closely applied to the entire surface of each chromosome is in my 
opinion beyond much doubt. To make an actual demonstration of the plasmatic 
membrane during its close application to the surface of the chromosomes is obviously 
out of the question. We are, therefore, obliged to accept less convincing evidence. 
In fig. 22, however, we have undoubtedly the beginning of such a process. If we 
compare figs. 22 and 23, and examine them in the light of the evidence obtained 
in similar stages in the case of Disporum, there is really only one rational conclusion to 
come to. It is, namely, that each chromosome has not only been completely and closely 
enveloped by a membrane, but each of these bodies, as a result of that enveloping 
process, has been furnished with a system of kinoplasmic fibrils. With the establish- 
ment of as many osmotic systems as there are chromosomes — systems which are to a 
great extent quite independent of one another— there will naturally follow a new and 
rapid readjustment of the lines of tension expressed in the kinoplasm. Such a 
readjustment is probably taking place in the stage represented in fig. 24. 

As stated above, the chromosomes in Gladiolus are very long, being many times 
longer than broad. It would therefore follow that as the nuclear membrane became 
closely applied to the chromosomes, the broad sides of these bodies, offering the greater 
osmotic surfaces, would naturally be thus furnished with many more kinoplasmic threads 
than the short end surfaces. This condition may be seen quite clearly in figs. 23 and 
24. And so with the final readjustment, the main lines of kinoplasm would find 
themselves exerting a tension on opposite sides and at right angles to the long axis 
of each chromosome, and these latter bodies would thus be suspended at the equator, 
as shown in fig. 25. 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 151 

Yucca. 

A perfect series of stages was obtained in the microspore mother-cells of Yucca, 
which furnished convincing evidence that the nuclear membrane neither breaks down 
nor collapses during the period of spindle formation. In fig. 26 we have represented a 
median section of a mother-cell with the bivalent chromosomes just about completely 
organised. The nuclear vacuole has reached its maximum size and occupies a space 
nearly equal to the volume occupied by the cytoplasm. That the nuclear vacuole is 
under high osmotic pressure seems evident from its spherical form and the distended 
condition of the enveloping membrane. The cytoplasm is finely and densely granular, 
but perfectly uniform in its reticulate structure. The peripheral cytoplasm is in 
close touch with the inner surface of the thick cell-wall. 

Fig. 27 represents a similar section at a somewhat later stage. It will at once be 
seen that the nuclear vacuole is now less than half the original volume shown in fig. 26. 
It will also be seen that this reduction is accompanied by a differentiation or rather a 
changed configuration of the cytoplasm. With the nuclear membrane as a base there 
appear numerous long radiating threads or fibrils which show every indication of having 
been drawn out from the cytoplasmic reticulum. Many of these threads are so fine that 
they can only be seen with difficulty, being more or less obscured by food granules. 
Others, again, are well defined and extend for some distance towards the periphery. In 
the stage shown in fig. 28 we find the nuclear vacuole has diminished still further and 
the kinoplasmic threads have increased in number. Fig. 29 represents a little later 
stage, where the nuclear vacuole has been reduced to such an extent that the chromo- 
somes have become crowded together by the enclosing nuclear membrane. The 
kinoplasmic threads are still more numerous and much more sharply defined. The lines 
of tension represented in the threads or fibrils seem to shift when the stage represented 
in fig. 30 has been reached. As indicated in this figure, the threads appear to arrange 
themselves in conical-shaped sheaves or groups, until, as shown in fig. 31, there is a 
distinct tendency to the multipolar arrangement. It should also be noted in this last 
figure that the nuclear membrane is now in close touch with the majority of the 
chromosomes. 

A comparative study of these serial stages (figs. 26 to 31) establishes a number of 
facts in regard to spindle development that are of vital importance. In the first 
place, it becomes obvious that there has been a gradual and progressive diminution in 
the amount of karyolymph. In the second place, there is no doubt whatever that the 
nuclear membrane persists throughout all of these stages. In the third place, these 
figures demonstrate quite clearly that the differentiation of the cytoplasmic reticulum 
into the kinoplasmic threads or fibrils progresses with the diminution in the volume of 
the nuclear vacuole. 

I have already stated my interpretation of these facts in the case of Disporum and 
Gladiolus. The diminishing of the karyolymph I believe to be due to osmotic 



152 MR A ANSTRUTHER LAWSON ON 

diffusion, and the presence of the plasmatic nuclear membrane makes such a 
diffusion possible. As this gradual diffusion progresses and the volume of the nuclear 
vacuole decreases, the reticulate cytoplasm finds itself under tension in being obliged 
to occupy a greatly increased cubical space (compare fig. 31 with fig. 26). This 
tension finds an expression in a changed configuration of the reticulum. This latter 
becomes drawn out in the form of threads by the diminishing and receding plasmatic 
membrane. 

A comparison of such stages represented in figs. 31 and 32 convinces me that the 
reduction of the karyolymph by diffusion does not cease when it becomes no longer 
visible as a clear nuclear sap, but continues until each chromosome becomes closely 
enveloped by the nuclear membrane. It will be seen from fig. 32 that each 
chromosome is not only thus furnished with a system of fibrils which become closely 
applied to its surface, but osmotic diffusion continues for some time, establishing new 
lines of tension. As in the case of Disporum and Gladiolus, the main lines of kino- 
plasm extend from the broad sides of the chromosomes, and the tension thus becomes 
exerted on opposite sides, and at right angles to the long axis of each bivalent 
chromosome. These latter bodies thus find themselves suspended at the equator, as 
shown in fig. 33. 

Hedera. 

With the object of extending the range of my observations into the Dicotyledons, 
I have selected the common ivy — Hedera helix — as a type for study. In the microspore 
mother- cells of this plant I find also a series of stages showing the persistence of the 
nuclear membrane throughout the prophase — stages which can only be interpreted in 
the manner for the monocotyledonous types, Disporum, Gladiolus, and Yucca. The 
nucleus, at the time the chromosomes are nearly formed from the spireme, occupies more 
than half the volume of the cell-cavity, as shown in fig. 34. In section the cytoplasm 
appears as a narrow zone filling the space between the nuclear membrane and the cell- 
wall. Its reticulate structure is uniform throughout. The chromosomes are small, oval- 
shaped bodies, and for the most part occupy a position in touch with the nuclear 
membrane. The large nuclear vacuole is evidently under high osmotic pressure. It is 
almost spherical in form and in a state of turgescence. 

Fig. 35 represents a similar section at a somewhat later stage. It will be seen that 
the nuclear vacuole is much smaller than that shown in fig. 34. A certain amount of 
the karyolymph has evidently diffused into the cytoplasm, and this has been carried 
still further in the stage shown in fig. 36. The conditions shown in figs. 35 and 36 are 
quite similar to corresponding stages in the mother-cells of Disporum. Accompanying 
the decrease in volume of the nuclear vacuole there is a change in the form of the 
cytoplasmic reticulum. It will be seen (fig. 35) that the reticulum, at one side of the 
nucleus, lias been drawn out into a broad tuft of short threads or fibrils. It will also 
be seen that as the nucleus becomes smaller and smaller (fig. 36), the kinoplasmic 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 153 

threads become longer and more radial in their distribution ; they also become more 
numerous and more sharply defined. 

My interpretation of these kinoplasmic threads in Hedera is the same as that for the 
types mentioned above. These threads simply express lines of tension in the cyto- 
plasm — a tension caused by the increased cytoplasmic area, and which is sufficient to 
change the configuration of the reticulum, and the threads are therefore the passive 
results of nuclear osmotic changes. 

In fig. 37 we have a typical multipolar stage with the numerous chromosomes 
crowded together in the small remaining vacuole of karyolymph. Nearly all of the 
chromosomes are now in close touch with the nuclear membrane. The closing in of 
the membrane continues until it completely envelops each separate bivalent chromo- 
some in the manner above described for Disporum. The large number of chromosomes in 
Hedera, however, complicates matters, and we have the confused appearance represented 
in fig. 38. I have not the slightest doubt, however, that the fibrils become attached to the 
individual chromosomes in this manner. With the establishment of so many smaller 
osmotic systems there would naturally follow a final readjustment of the lines of tension 
expressed in the achromatic figure. This readjustment would be mainly controlled by 
the osmotic surfaces enveloping each chromosome. There was no indication whatever 
to suggest the bodily shifting and ultimate coalescence of the sheaves of the multi- 
polar figure. The establishment of the equatorial plate and the symmetrical bipolar 
arrangement of the lines of tension represented in the fibrils is evidently brought about 
in the manner already described above in the case of Disporum and Gladiolus. 



Allium. 

It is well known that spindle development in somatic cells is somewhat different 
from that occurring in spore mother-cells. This fact was established by N£mec (1899) 
and others, who found that as the kinoplasmic weft becomes differentiated about the 
nuclear membrane, it takes the form of two conical-shaped caps which project from 
opposite sides of the nuclear cavity. These kinoplasmic projections, on account of 
their position, are commonly referred to as polar caps. The threads of which they 
are composed eventually become the main fibrils, and their apices become the poles of 
the spindle. So that in vegetative cells there is nothing in the nature of the multipolar 
arrangement which is so characteristic of the heterotype mitosis. It seems that the 
mitotic spindle in somatic cells is always bipolar from the beginning. 

Because of this striking difference between the somatic and heterotype mitosis, 
and in view of the new facts and interpretations recorded above in the cases of 
Disporum, Gladiolus, Yucca, and Hedera, I have considered it advisable to re-examine 
the nuclear changes leading to the formation of the achromatic spindle in the root tip 
of Allium cepa, a plant which has become a classic for nuclear study. 

The first indication of approaching mitosis is a decided enlargement of the nuclear 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 7). 24 



154 MR A. ANSTRUTHER LAWSON ON 

vacuole in which the chromatin reticulum lies suspended in the karyolymph. This 
enlargement is not so great as that which occurs at a similar time in the heterotype 
mitosis, but nevertheless sufficient to make these nuclei very conspicuous. The 
chromatin now changes from the finely divided reticulate condition to long, rather 
stout threads, and ultimately into the compact definite chromosomes, as shown in 
fig. 40. This figure represents a section of a somatic cell taken at a time when the 
chromosomes are nearly organised. The nuclear vacuole, it will be seen, is perfectly 
spherical and obviously under high osmotic pressure. It should be noted in passing 
that, unlike sporogenous cells, there are several large vacuoles in the cytoplasm. 

In the following stages, namely, those represented in figs. 41, 42, 43, and 44, I was 
able to confirm the main observations of Nemec (1899) and others in regard to the 
development of the two polar caps at opposite sides of the nucleus. The beginning 
of these kinoplasmic caps is shown in fig. 41, where they appear in section as 
shallow crescent-shaped groups of threads. As indicated in figs. 42, 43, and 44, these 
crescent-shaped caps appear to elongate and become decidedly conical in form. These 
features do not differ essentially from what is commonly known in this connection. 
Unfortunately, however, a very interesting and important fact has been overlooked. 
It is, namely, that as these kinoplasmic structures known as the polar caps develop, 
there is a corresponding diminution in the volume of the nuclear vacuole. This is 
so clearly illustrated in figs. 41, 42, 43, 44, and 45, that actual measurements 
are not necessary. The decrease in the amount of karyolymph is gradual but 
unmistakable. It will also be noticed from these figures that as the karyolymph 
diminishes the outline of the nucleus becomes less spherical. As shown in figs. 43, 44, 
and 45, the nuclear vacuole becomes decidedly flattened on the sides that form the 
bases of the polar caps. 

There is no doubt whatever in my mind that the factors responsible for the forma- 
tion of the polar caps are the same as those which we have described above as being 
concerned in the formation of the kinoplasm during the heterotype mitosis. They are, 
namely, that a state of tension has been created in the cytoplasm by the reduction 
in the volume of the nuclear vacuole, and the cytoplasm thus finds itself obliged to 
occupy a greatly increased cubical space. This tension finds an expression in the 
drawing out of threads from the cytoplasmic reticulum by the receding nuclear 
membrane. 

It is curious that somatic mitosis should be characterised by the drawing out of 
only two conical-shaped sheaves of kinoplasm. This feature, however, becomes less 
difficult to understand when we remember that the numerous vacuoles which are 
always present in the cytoplasm would render a radial or a multipolar arrangement 
of the kinoplasm impossible. In sporogenous cells these cytoplasmic vacuoles do 
not occur. 

The developmental stages that now follow are practically the same as those 
described above for the heterotype spindle. The diffusion of the karyolymph continues 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 155 

until the nuclear membrane completely envelops each individual chromosome (figs. 46 
and 47), thus providing each of these bodies with a series of fibrils. On account of 
the great length of the chromosomes, the greater number of fibrils — and consequently 
the dominating lines of tension— would be drawn to the broad sides of the chromosomes, 
and in this fashion the equatorial plate would be established, and the mature spindle 
presents the appearance shown in fig. 48. 

An examination of fig. 48 will convince one that the poles of the spindle are not 
very remotely situated from the equator. It will also be seen without any doubt in 
figs. 49 and 50 that the daughter chromosomes move in opposite directions beyond the 
positions of the actual poles of the spindle. From this fact alone it would seem that 
the attached fibrils are not actively concerned in this movement. As I have stated 
elsewhere, they may possibly act as guide lines. 

The organisation of the daughter nuclei, as illustrated in figs. 50, 51, 52, and 53, is 
important and interesting, because it shows how the tension originally set up in the 
cytoplasm during the prophase and expressed in the spindle fibrils eventually becomes 
relaxed. In fig. 50, for instance, we see the daughter chromosomes arranged in two 
groups at opposite sides of the cell. The state of tension in the cytoplasm is still 
evident by the drawn-out threads of cytoplasm stretching between the two groups. 
This same condition is still shown in fig. 51, where the chromosomes in each group are 
closely massed together. In fig. 52 we have a stage where each chromosome has 
become more or less vacuolated, clearly indicating that a considerable amount of 
karyolymph has been taken in by endosmosis. This would necessarily release the 
tension in the cytoplasm to a considerable extent. When, finally, as shown in fig. 53, 
the combined volumes of the two daughter nuclei approximates the original mother 
nucleus, all tension is relaxed and all fibrils disappear. 

The interpretations which I have given above for these new stages in spindle 
formation in the various types mentioned are quite in harmony with certain well-known 
facts concerning important changes which take place in the nucleus both before and 
after mitosis. In the first place, it is an established fact that in the re-organisation of 
the daughter nuclei there is an accumulation of karyolymph within the chromosomes. 
This nuclear sap first appears in the form of minute lacuna? which increase in size and 
flow together, giving the chromosomes a vacuolated appearance. With the accumula- 
tion of the karyolymph the chromosomes become very finely divided, and the chromatin 
eventually appears as a fine reticulum suspended in a large vacuole of nuclear sap. 
It is also well known that out of this finely divided condition there is developed 
a stage commonly called the spireme, where the chromatin assumes the form of long, 
fairly thick threads, and not so finely divided as in the reticulum. And, finally, the 
spireme threads give rise to the more compact chromosomes. Now, the point of 
interest is that during this transition period between the finely divided condition of the 
reticulum and the more compact condition of the chromosomes there always occurs 



156 MR A. ANSTRUTHER LAWSON ON 

an enlargement of the nuclear vacuole, that during this period there is an increase in 
the amount of contained karyolymph. In spore mother-cells this enlargement is very 
great indeed, and constitutes a conspicuous growth period (Lawson, 1911), while in 
somatic cells it is not so great. Now, coupling these facts with the building up of new 
chromatin substance with each nuclear generation, there is a suggestion that the 
condition of the chromatin has an influence on the varying osmotic relations. For 
now, again, it would seem that as soon as the chromatin assumes the compact form of 
the chromosomes there immediately follows a diminution in the volume of the 
karyolymph, and this continues until the nuclear sap is completely diffused into the 
cytoplasm. 

This investigation, however, is not intended to cover all the phenomena of mitosis ; 
its only object has been to throw some light on the problem of the achromatic spindle 
and the factors responsible for its formation in vascular plants. 

The writer wishes to express his indebtedness to the Executive Committee of the 
Carnegie Trust for a grant to defray the cost of the plates illustrating this paper. 

Summary. 

A study of the microspore mother-cells of Disporum, Gladiolus, Yucca, Hedera, 
and the vegetative cells in the root tip of Allium has revealed a series of stages in the 
development of the mitotic spindle which have never before been described. 

The new stages that have been discovered are to be found in the prophase immedi- 
ately preceding the organisation of the equatorial plate, and concern the fate of the 
nuclear membrane. 

The interpretation of these stages has thrown a new light on the process of mitosis, 
and necessitates a revision of the accepted views of nuclear phenomena. 

Contrary to the generally accepted view, it has been found that the nuclear mem- 
brane does not break down or collapse at any period during spindle development, but 
behaves as one would expect a permeable plasmatic membrane to behave under varying 
osmotic relations. 

The nucleus is regarded as an osmotic system, and its membrane constitutes an 
essential element in that system. 

It is a fact of common knowledge that the chromatin changes both in quantity and 
form sometime before the metaphase. The chromatin must increase in quantity, because 
the same amount is present for each mitosis. It changes in form from the finely 
divided condition represented in the reticulum and spireme to the more compact and 
homogeneous form of the chromosomes. 

It would seem that these changes are in some way responsible for a variation in the 
osmotic relations of the karyolymph : at any rate, a gradual diffusion of the nuclear 
sap immediately follows these changes in the form of the chromatin. 

A series of stages has been found showing beyond doubt that, closely following the 
organisation of the bivalent chromosomes, there is a gradual diminution in the volume 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 157 

of the nuclear vacuole. It is believed that the karyolymph gradually diffuses by 
exosmosis into the cytoplasm. 

Throughout the entire prophase the nuclear membrane is functional in this osmotic 
transfer. 

As the nuclear vacuole becomes smaller and smaller, the membrane gradually 
closes in about the chromosomes. These latter become crowded together about the 
nucleolus. 

When the karyolymph becomes so much reduced that it is no longer visible as a 
clear nuclear sap, the membrane becomes closely applied to, and completely envelops, 
the surface of each chromosome. 

The result is that, instead of a single osmotic system represented in the nucleus, we 
now have established as many independent osmotic systems as there are chromosomes. 

For some time previous to the diffusion of the karyolymph the nuclear vacuole 
occupies a space that may approach or even exceed half the volume of the cell- 
cavity. So that all of these circumstances bring about a condition where a limited 
amount of cytoplasm of reticulate structure is obliged to occupy a cubical space which 
has greatly increased by the reduction in the volume of the nuclear vacuole. 

This necessarily sets up a tension in the cell — a tension sufficient to cause a re- 
adjustment and changed configuration in the reticulate form of the cytoplasm. 

As the nuclear vacuole becomes smaller and smaller, the cytoplasm in the region of 
the nuclear membrane becomes changed to the form of fine threads or fibrils which are 
drawn out from the reticulum by the receding membrane. 

The state of tension set up in the cytoplasm thus finds an expression in these drawn- 
out threads of " kinoplasm." 

From the different plants studied it seems that the lines of tension as expressed 
in the fibrils may group themselves in various ways at first. Thus we may have a 
weft of kinoplasm surrounding the nucleus ; or a system of kinoplasmic radiations ; or, 
more commonly, a number of conical-shaped sheaves of fibrils. 

But whichever form the kinoplasm may appear to take, the lines of tension are 
constantly shifting throughout the prophase. 

Such a shifting does not mean the changing of the threads bodily from one position 
to another. It means the relaxing of the tension along certain threads, which would 
consequently fall back into the reticulum form, and the setting up of new lines of 
tension by the drawing out of threads from the hitherto undifferentiated reticulum. 

In this fashion not only individual threads, but entire sheaves or cones of fibrils 
may appear to assume different positions. 

For these and other reasons the generally accepted view that the sheaves or cones 
of fibrils in the multipolar figure approach one another and eventually coalesce into 
two groups should no longer be retained. 

There was no evidence to support the generally accepted view that the spindle 
fibrils grow into the nuclear area and attach themselves to the chromosomes. 



158 MR A. ANSTRUTHER LAWSON ON 

This attachment is undoubtedly brought about by the enveloping of each bivalent 
chromosome by the receding nuclear membrane. This membrane, which is really the 
base of the drawn-out threads, upon being closely applied to the surface of each bivalent 
chromosome, furnishes each of these bodies with its own system of kinoplasm. 

No evidence was found to support the view that the contraction of the attached 
fibrils draws the daughter chromosomes to the poles of the spindle. Such fibrils may 
serve as guide lines, but take no active part in the movement. 

Taking all of the facts into consideration, I am forced to the conclusion that 
the achromatic spindle in vascular plants is simply an expression of a state of 
tension in the cytoplasm, and that this tension is caused in the first place by nuclear 
osmotic changes which create a condition where a limited amount of cytoplasm is 
obliged to occupy an increased cubical space. 

As a result of this investigation, I can no longer regard the achromatic figure as 
an active factor in mitosis. It seems to be nothing more than a passive effect of 
nuclear osmotic changes. 



LITERATURE CITED. 



Allen, C. E., 1903, " The Early Stages in Spindle-Formation in the Pollen Mother-cells of Larix." Ann. But., 

vol. xvii. p. 281, 1903. 
1905, "Nuclear Division in the Pollen Mother-cells of Lilium Canadense." Ann. Bot., vol. xix. 

p. 189, 1905. 

1905, "Das Verhalten der Kernsubstanzen der Synapsis in den Pollenmutterzellen von Lilium 



Canadense." Jahrb. Wiss. Bot., xlii. p. 72, 1905. 
Belajeff, W., 1894, "Zur Kenntniss der Karyokinese den Pflanzen." Flora, lxxix., p. 430, 1894. 
Berghs, J., 1904, " La formation des chromosomes he^erotypiques dans la sporogenese vegetale." La Cellule, 

xxii. p. 43, 1904. 
■ 1905, "La formation des chromosomes heterotypiques dans la sporogenese vegetale." La Cellule, 

xxii. p. 141, 1905. 

1905, " Le fuseau heterotypique de Paris quadrafolia." La Cellule, xxii. p. 203, 1905. 



Byxbee, E. S., 1900, "The Development of the Karyokinetic Spindle in Lavatera." Proc. Gal. Acad., Hi., 

Bot. ii. p. 63, 1900. 
Cardiff, J. D., 1906, "A Study of Synapsis and Reduction." Bull. Torr. Bot. Club, xxxiii. p. 271, 1906. 
Davis, B. M., 1899, "The Spore Mother-cells of Anthoceros." Bot. Gaz. 

1909, "Pollen Development of Oenothera grandiflora." Ann. Bot., xxiii. p. 551, 1909. 

"The Reduction Division of CEnotliera biennis." Ann. Bot., xxiv. p. 631, 1910. 

Debski, B., 1897, " Beobachtungen iiber Kerntheilung bei Chara fragilis." Jahrb. Wiss, Bot., xxx. p. 227, 

1897. 
Dixon, H. H., 1901, "On the First Mitosis of the Spore Mother-cells of Lilium." Notes Bot.Sch. Trin. Coll. 

Dublin, 1901. 
Farmer. J. B., 1893, " On Nuclear Division in the Pollen Mother-cells of Lilium martagon." Ann. Bot., vii. 

p. 392, 1893. 
1895, "Ueber Kerntheilung in Lilium, Antheren, besonders in Bezug auf die Centrosomenfrage." 

Flora, lxxx. p. 56, 1895. 
and Moore, 1905, "On the Maiotic Phase in Animals and Plants." Quar. Jour. Micro. Soc, xlviii. 

p. 489, 1905. 
and Dioby, L., 1910, " On the Cytological Features exhibited by certain Varietal and Hybrid Ferns." 

Ann. Bot. xxiv., p. 191, 1910. 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 159 

Gates, R. R., 1909, "The Behaviour of Chromosome in (Enothera lata x 0. gigas." Bot. Gaz., xlviii. 
p. 179, 1909. 

1908, "A Study of Reduction in (Enothera rubrinervis." Bot. Gaz., xliii. p. 1, 1908. 

Gregoire, U., et Wygaerts, 1903, "La reconstitution du noyau et la formation des chromosomes dans les 

cinises somatiques." La Cellule, t. xxi., 1903. 
Hartog, M., 1905, " The Dual Force of the Dividing Cell." Pt. I. The achromatic spindle figure illustrated 

by magnetic chains of force, Proc. Roy. Soe. B., lxxvi. p. 548, 1905. 
Juel, O., 1897, "Die kerntheilungen in den Pollenmutterzellen von Hemerocallis fulva, und die bei 

denselben auftreten den Unregelmassigkeiten." Jahrb. Wiss. Bot., xxx. p. 205, 1897. 
Lawson, A. A., 1900, "Origin of the Cones of the Multipolar Spindle in Gladiolus." Bot. Gaz., xxx. 

p. 145, 1900. 
1903, "On the Relationship of the Nuclear Membrane to the Protoplast." Bot. Gaz., xxxv. p. 305, 

1903. 

1903, "Studies in Spindle Formation." Bot. Gaz., xxxvi. p. 81, 1903. 

1911, "The Phase of the Nucleus known as Synapsis." Trans. Roy. Soc. Edin., vol. xlvii., pt. iii. 



p. 591, 1911. 
Miyake, K., 1905, "Ueber Reduktionsteilung in den Pollenmutterzellen einigen Monokotylen." Jahrb. Wiss 

Bot, xlii. p. 83, 1905. 
Moore, A. C, 1903, "The Mitosis in the Spore Mother-cell of Pallavicinia." Bot. Gaz., xxxvi. p. 384, 1903. 
Mottier, D. M., 1898, " Ueber das Verhalten der Kerne der Entwickelung des Embryosacks und die Vorgange 

bei der Befruchtung." Jahrb. Wiss. Bot., xxxi. p. 125, 1898. 
1897, " Beitrage zur Kenntniss der Kerntheilung in den Pollenmutterzellen einiger Monokotylen und 

Dikotylen." Jahrb. Wiss. Bot., xxx. p. 169, 1897. 

1907, "The Development of the Heterotypic Chromosomes in Pollen Mother-cells." Ann. Bot., xxi. 



p. 309, 1907. 
Nkmec, B., 1899, "Ueber die karyokinetische Kerntheilung in der Wurzelspitze von Allium cepa." Jahrb. 

Wiss. Bot., xxxiii. p. 313, 1899. 
Osterhout, W. J. V., 1897, "Ueber Entstehung der karyokinetischen Spindel bei Kquisetum." Jahrb. 

Wiss. Bot., xxx. p. 159, 1897. 

■ 1902, "Spindle Formation in Agave." Proc. Cal. Acad. Sci., iii., Bot. ii. p. 255, 1902. 

Overton, J. B., 1909, "On the Organisation of the Nuclei in the Pollen Mother-cells of Certain Plants." 

Ann. Bot., xxiii., 1909. 
Peirce, G. J., Plant Physiology. New iTork, 1903. 
Pfeffer, W., 1890, Zur Kenntniss der Plasmahaut und der Vacuolen, etc. Leipzig, 1890. 

1897, Pjianzenphysiologie. Leipzig, 1897. 

1900, Physiology of Plants. Translated by Ewart, Clarendon Press, Oxford, 1900. 

Sargant, E., 1897, "The Formation of the Sexual Nuclei in Lilium Martayon." Ann. Bot., xi., 1, 1897. 
Smith, R. W., 1900, "The Achromatic Spindle in the Spore Mother-cells of Osmunda regalis." Bot. Gaz., 

xxx. p. 361, 1900. 
Stomps, T. J., 1910, Kerndeeling en Synapsis bij Spinacia Oleracea. Amsterdam, 1910. 
Strasburger, E., 1882, "Ueber den Teilungsvorgang der Zellkerne und das Verhaltniss der Kerntheilung 

zur Zelltheilung." Arch. Mikr. Anat., xxi. p. 476, 1882. 
1900, " Ueber Reduktionsteilung, Spindelbildung, Centrosomen und Cilienbilder im Pflanzenreich " 

Hist. Beitr., vi. p. 224, 1900. 
1905, "Typische und allotypische Kernteilung." Jahrb. Wiss. But., xlii. p. 1, 1905. 

1907, "Ueber die Individuleitat der chromosomen und die Pfropfhybriden-Frage." Jahrb. Wiss. 



Bot., xliv. p. 482, 1907. 
Williams, C. L., 1899, "The Origin of the Karyokinetic Spindle in Passiflora coerulea." Proc. Cal. Acad. 

Sci., iii., Bot. i. p. 189, 1899. 
Yamanouchi, S., 1906, "The Life History of Polysiphonia." Bot. Gaz., xlii. p. 401, 1906. 

1908, " Sporogenesis in Nephrodium." Bot. Gaz., xlv. p. 1, 1908. 

1910, "Chromosomes in Osmunda." Bot. Gaz., xlix. p. 1, 1910. 



160 MR A. ANSTRUTHER LAWSON ON 

EXPLANATION OF FIGURES. 

All the figures were drawn with the aid of the camera lucida, with Zeiss compensating oculars and oil 
immersion objective one-twelfth at the magnifications indicated. 
Figures 1 to 13 x 1900. 
Figures 14 to 25x1400. 
Figures 26 to 53x1900. 

DlSPORUM. 

Fig. 1. A section of a microspore mother-cell when the nuclear cavity has reached its maximum size. Five 
bivalent chromosomes are to be seen and a large nucleolus. The cytoplasmic reticulum shows a uniformity 
of structure. 

Fig. 2. The same at a later stage. A considerable diminution in the size of the nuclear cavity is to be 
seen. The cytoplasm in the vicinity of the nuclear membrane has lost its reticulate structure and takes the 
form of a series of delicate threads which appear to radiate from the membrane. 

Fig. 3. The same at a still later stage, with a further diminution in the size of the nuclear cavity. The 
cytoplasmic radiations have become longer and more sharply defined. 

Fig. 4. The same, showing that the nuclear cavity has diminished to less than half the cubical area shown 
in fig. 1. The membrane is still intact, but a great reduction in the amount of karyolymph has taken place. 
The kinoplasmic radiations are more sharply defined. 

Fig. 5. Another cell showing about the same conditions, but with the nucleus more centrally situated. 

Fig. 6. A still later stage of the same, showing the crowding together of the chromosomes as the nuclear 
membrane closes in about them. 

Fig. 7. The amount of karyolymph has been so much reduced that the nuclear membrane is in close touch 
with the chromosomes for the greater part of its surface. 

Fig. 8. The karyolymph can no longer be seen as a clear nuclear sap, and the nuclear membrane has 
enveloped each individual chromosome. 

Fig. 9. As the nuclear membrane has enveloped each bivalent chromosome, each of these bodies becomes 
provided with a sheaf of fibrils which to all intents and purposes are attached. As these sheaves of fibrils 
appear on opposite sides, there is a small spindle for each bivalent chromosome. 

Fig. 10. The metaphase with the daughter chromosomes separating, and have begun their movement to the 
poles of the spindle. 

Fig. 11. The daughter chromosomes at the poles. 

Fig. 12. Showing the vacuolisation of the daughter chromosomes by the accumulation of karyolymph and 
the organisation of a membrane about each daughter nucleus. 

Fig. 13. The daughter nuclei completely organised. 

Gladiolus. 

Fig. 14. A section of a microspore mother-cell with the nucleus at its maximum size. 

Fig. 15. A slight diminution in the size of the nuclear cavity is shown, and a narrow weft of kinoplasmic 
threads appears about the nuclear membrane. This weft is not uniform, being much more prominent at 
intervals. 

Fig. 16. A typical multipolar condition of the kinoplasm. 

Fig. 17. The same a little later, showing four cones in the multipolar condition. The nuclear vacuole 
has decreased to less than half its original dimension. 

Fig. 18. The same, showing three cones in the section. 

Fig. 19. Another example of the same stage, showing two cones. 

Fig. 20. A little later stage, showing the nuclear membrane still intact. 

Fig. 21. Another example about the same stage. 

Fig. 22. The nuclear vacuole has so much reduced that the membrane is seen closing in about the 
chromosomes. 

Fig. 23. The nuclear membrane has now enveloped each chromosome. 

Fig. 24. The same a little later. 

Fig. 25. The mature spindle with the chromosomes at the equator. 



NUCLEAR OSMOSIS AS A FACTOR IN MITOSIS. 161 



Yucca. 

tig. 26. A median section of a microspore mother-cell, showing the nuclear vacuole at its maximum size. 

Fig. 27. The same, showing a considerable reduction in the volume of the nuclear vacuole and an 
indication of radiating kinoplasmic threads in the surrounding cytoplasm. 

Fig. 28. The same at a slightly later stage, with the kinoplasmic threads much more sharply defined. 
The nuclear vacuole is considerably less than a fourth of its original volume, and the nuclear wall is still 
intact. 

Fig. 29. Another example of the same condition, but slightly older. 

Fig. 30. The nuclear membrane still intact, and the kinoplasmic threads appear in groups or sheaves. 

Fig. 31. The chromosomes are crowded together by the enclosing of the nuclear membrane about them. 
An irregular multipolar condition has been reached. 

Fig. 32. The nuclear membrane has completely enveloped each chromosome, so that each of these bodies 
is now provided with a system of kinoplasmic threads. 

Fig. 33. The mature spindle with the chromosomes at the equator. 

Hedera. 

Fig. 34. A median section of a microspore mother-cell showing the nuclear vacuole at its maximum'size. 

Fig. 35. The same, showing a considerable decrease in the volume of the nuclear vacuole and the 
appearance of an incomplete zone of kinoplasm. 

Fig. 36. The same, showing the kinoplasmic threads radiating from the nuclear membrane and much 
more sharply defined. There is a still further decrease in the nuclear vacuole, and the nuclear membrane 
is intact. 

Fig. 37. A distinct but irregular multipolar condition. The karyolymph in the nuclear vacuole has 
almost completely diffused, and the chromosomes are consequently crowded together by the receding nuclear 
membrane. 

Fig. 38. Each chromosome has become enveloped by the nuclear membrane and consequently furnished 
with a system of kinoplasmic fibrils. 

Fig. 39. The mature spindle with the chromosomes at the equator. 

Allium. 

Fig. 40. A median section of a vegetative cell in the root tip. The nucleus is quite large and prepared 
for mitosis. The chromosomes are almost organised. 

Fig. 41. A slightly later stage of the same. The kinoplasm appears in the section as narrow crescents 
of threads which appear at opposite sides of the nucleus. There has been a slight decrease in the size of 
the nucleus. 

Fig. 42. A further decrease in the volume of the nucleus, and a corresponding increase in the 
differentiation of the kinoplasm. 

Fig. 43. The same a little later, showing the polar caps. 

Fig. 44. Another section of the same. 

Fig. 45. The polar caps completely organised, with a very obvious diminution in the volume of the 
nuclear vacuole. The nuclear membrane is still intact. 

Fig. 46. The nuclear membrane has closed in about each chromosome. 

Fig. 47. Showing the connection of the kinoplasmic fibrils to the membrane enveloping each chromosome. 

Fig. 48. The mature spindle with the chromosomes at the equator. 

Fig. 49. The daughter chromosomes have moved to the poles of the spindle. Numerous threads of 
kinoplasm stretch between the chromosomes at opposite ends of the spindle. 

Fig. 50. The grouping together of the daughter chromosomes at the poles. 

Fig. 51. The same a little later. 

Fig. 52. The vacuolization of the daughter chromosomes, accompanied by a relaxation of the tension in 
the cell. This relaxation is expressed in the loose curvature of the kinoplasmic threads. 

Fig. 53. The daughter nuclei fully developed. The development of the large nuclear vacuoles has 
relaxed the tension in the cytoplasm completely, and all kinoplasmic threads have vanished. 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 7). 25 



.Roy Soc Edxn T 



LAWSON- NUCLEAR OSMOSIS. PI I 



Vol. xlvhi. 






/ 









■•(. 




c 













•'- 


















m 



■ ■ 

i. 


















d 



m 



8. 









10. 













1 -L/ 






11. 



12. 



13. 



Hath,Lith T London 



/^H~rv,0', 



4"iAL ^ 



5. Roy. Soc.Edin 1 



LAW SON — NUCLEAR O S M O S I S .Pl.H Vol XLVIII. 




.del. 



Hntb., lilt 1 * London 




S>* MUi 






-\ay. Soc.Edin. r 



LAWSON .— NUCLEAR OS M O S I S.Pl.ni. 



Vol. XL-VET 




38 



39. 



.n.aeL 



HutkLitV London,. 



^ : 



"r 



4-4. 



'3, I 




4* 





! 7""7'' 




V .-, 


■1 


U'v-' 








LAWSON- NUCLEAR O S M S 1 S.Pl.IV: 



Voi.nvm. 



J 




51. 




4-6. 



47. 



50. 






'■■'■■> ■•-: 



^'a~ 






53. 




a?.i. 



Butli Litk^" london . 



(163) 



VIII. — On the Structure and Affinities of Metaclepsydropsis duplex (Williamson). 
By W. T. Gordon, M.A. ; B.A., D.Sc., Falconer Fellow of Edinburgh University, 
Lecturer in Palaeontology, Edinburgh University. Communicated by Professor 
James Geikie, D.C.L., LL.D., etc. (With Four Plates.) 

(Read July 3, 1911. MS. received July 28, 1911. Issued separately December 18, 1911.) 

Introduction. 

While collecting specimens of Diplolabis romeri (Solms) at Pettycur about three 
years ago, I discovered several blocks of stone containing numerous petrified frag- 
ments of another Zygopterid fern — Metaclepsydropsis duplex (Williamson). Two of 
the blocks were found to be parts of one large mass, and the petrifactions could be 
traced from the one block into the other. In size the complete mass must have been 
about 3 feet x 2 feet x 2 feet. The larger portion had to be broken into two before it 
was possible to remove it. Another block containing similar specimens was obtained 
later, but the preservation was so poor that the whole of the material was discarded. 

The Pettycur plants are usually preserved by an infiltration of calcareous material, 
chiefly carbonate of lime, but in these blocks the petrifying material was siliceous. 
This silica was present in two forms — chalcedony and crystalline quartz. (As a result, 
hollow stems had the appearance of agates, generally with a crystalline centre.) The 
exterior had been more or less weathered, and the specimens stood out on the surface, 
their tissues being perfectly visible, and giving a lace-like appearance to the surface of 
the blocks. 

As the occurrence of Diplolabis romeri was a new record from a British source, and 
as the specimens could be more easily examined (since thin sections are more readily 
prepared from calcareous than from siliceous material), I decided to finish my work on 
that genus before proceeding with a systematic investigation of the silicified specimens 
of Metaclepsydropsis duplex. 

On looking over the specimens, however, about a year after they had been collected, 
I noticed, among the petioles, a fragment of what appeared to be a fern stem. Two 
sections were prepared from it, and the length was ascertained by cutting that part of 
the block into pieces \ to f inch thick. The same stem was followed into another 
part of the block, and its extent in that direction also determined. The total length of 
this stem fragment was about 8 inches ; it was hollow in the centre, and throughout its 
length no petioles or roots were emitted. 

When this part of the block was cut up it was found to be so dark that no structure 
could be observed ; in the hope that exposure to the atmosphere would soon weather 
the surfaces and so render the specimens visible, another part was sawn into pieces 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 8). 26 



164 W. T. GORDON 

3 inches thick and the slabs placed out-of-doors for about nine months. Silica, however, 
takes a considerable time to weather, and even after all these months there was no 
appreciable difference. On experimenting with hydrofluoric acid I found that in a few 
minutes the surfaces could be etched, and after this treatment the petrifactions were 
even more clearly seen than when they had been naturally weathered, The alteration 
also was a mere surface one, and the specimens were not destroyed. So clearly did the 
specimens appear that the surface could be examined with a fairly high power — a 
tjr-incli objective — and even the thickenings on the walls of the tracheides could be 
made out by reflected light. It was thus possible to examine roughly the contents of 
a piece of material without making any thin sections. In this way much time was 
saved, and suitable examples of stems, petioles, and roots detected at once. 

As mentioned above, the Pettycur plants are usually preserved in calcareous material, 
and the thickenings on the cell walls are generally distinct. I have never, however, 
seen such perfect preservation as is exhibited in these silicitled specimens. Locally in 
the blocks the petrifying substance had been replaced by iron pyrites, and the tissue 
could only be seen by reflected light, when it appeared like black tracery in the yellow 
matrix. On the whole the preservation is magnificent. As previously remarked, 
however, the whole matrix is very dark, and so the preparations require to be very thin 
before they become transparent. With care the sections can easily be reduced to about 
•025 mm., the homogeneity of the material and its lack of cleavage rendering this easier 
than in the case where calcite is the petrifying medium. The test used to determine 
whether the sections were thin enough, was that employed in the preparation of rock 
sections, viz. silica (quartz) grey to clear between crossed nicols. 

In a jDrevious paper * it has been noted that the silica is present in two states, 
(a) chalcedony, and (b) quartz. The chalcedony appears to have been laid down first, and 
forms a layer round the wall of each element of the tissue. With this layer increasing 
in thickness the lumen is reduced in size. After the cell lumen is reduced to about one- 
quarter of its original size, the whole is filled in by crystalline quartz. All the silica, 
whether chalcedony or quartz, is almost perfectly transparent, and the strong contrast 
between the glassy interior of the cell and its dark walls has rendered the preparation 
of photo-micrographs much easier and more effective than in cases where calcite forms 
the matrix. In this latter case the cleavages sometimes interfere with the delicate cell 
structure, and so good photographs can hardly be obtained. 

In May of last year a more systematic study of the specimens contained in the 
fragments of the large block was begun. Several additional examples of the stem were 
obtained, and their occurrence among so many petioles might have been cited as a proof 
that stems and petioles belonged to one and the same plant. Such evidence, however, 
is of very little value and cannot be relied upon. This is especially the case when other 
genera occur in the same block. As a general rule the Pettycur blocks contain a great 
number of different genera ; indeed, practically the whole flora may be represented by 

* Gordon, Trans. Geul. Soc. Edin., vol. ix., 1909. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 165 

the specimens in one block. In the siliceous masses, on the other hand, there are 
singularly few genera, and at first sight this lends weight to the evidence of association. 
There was, however, another member of the Zygopteridese present — Stauropteris burnt- 
islandica, P. Bertrand, — and therefore the evidence of mere association must be dis- 
carded. The other genera present in the block were Botryopteris antiqua, Kidston ; 
Lepidophloios Scottii, Gordon ; Lepidostrobus cylindricus, Gordon in MS. ; Lepido- 
carpon Wildianum, Scott; Bensonites fusiformis, R. Scott; and Stigmaria jicoides, 
Brongn. 

In the meantime Miss Benson, D.Sc, F.L.S., had discovered a similar stem, also 
associated with numerous petioles of Metaclepsydropsis duplex, in one of the calcareous 
blocks from Petty cur. In her subsequent investigation Miss Benson obtained " a 
considerable amount of the stem (22 inches) without securing any well-preserved 
nodes." * A petiole-trace was closely associated with this stem towards one end, but, 
unfortunately, the stem emerged on the surface of the block just as the petiole-trace 
became fused with it. The specimen, which Miss Benson very kindly placed at my 
disposal and which is a co-type with my own specimens, illustrates clearly the great 
length of the internodes in this species, and proves the distance between two nodes, in 
some cases, to be at least 22 inches. 

Fragments of petioles of Metaclepsydropsis duplex were exceedingly numerous in 
the silicified blocks, but they were usually of no great length. Some were about 
8 inches long, but many did not exceed 3 inches, while others were under 1 inch in length. 
Most of these petioles were crushed, but often there were short lengths which showed 
no crushing. There were so many specimens that, by choosing the uncrushed portions 
only, a complete picture of the anatomy of the genus could be obtained. One remark- 
able feature about the specimens is that very few would be classed as normal petioles 
of M. duplex, and yet they would never be referred to any other genus, as will be 
seen later. 

By October 1910 sufficient positive evidence had been obtained to refer the stems 
found in the block to M. duplex. The more perfect specimens figured here were not 
discovered, however, until January 1911. Since then the early stages of petiolar 
development have been ascertained. The preparations examined number about 220. 
Most of these were prepared by myself, in order that nothing important might be 
lost during the process of preparation. As a rule the slides number 12 to 14 per 
inch of material, and they are generally in fairly long series. The series showing the 
branching of the stem and the departure of the petiole-trace was obtained from a piece 
of stem 3 inches long. From this piece 39 sections (including two longitudinal sections 
each | inch long) were prepared ; the sections were cut very close together, and run 
about 16 to the inch. 

Apart from the silicified specimens, I have made several series of sections from 
calcified examples as they happened to illustrate points not shown in the case of the 

* Letter, 17th May 1911. 



166 



W. T. GORDON 



siliceous petrifactions. For the sake of completeness, also, it is proposed to redescribe 
the species, particularly as previous accounts have been based on a few isolated sections 
from several distinct individual specimens, and thus several points have not been 
recorded. All previous work on this genus has been based on petioles whose xylem 
strand had an hour-glass shape, but I hope to show that this trace has been developed 
from a simpler type. 

Metaclepsydropsis duplex was first recorded by Williamson in 1874 under the 
name Rachiopteris duplex* His specimens had the hourglass-shaped xylem strand so 
characteristic of the mature petiole. From 1874 until P. Bertrand's description in 
1909 no further work was done on this genus, although it was constantly referred to. 
Kidston and Gwynne-Vaughan, in Part IV. of their memoir on the fossil Osmun- 
dace8e,t refer to Metaclepsydropsis, but have discussed its affinities rather than its 
anatomy. Apart from Kidston, no observer seems to have had more than a few isolated 
sections to deal with. Some years ago, however, Dr Kidston had a particularly fine series 
prepared, and these he has kindly placed at my disposal. In addition to showing the 
anatomy so far as it is known, they exhibit the remains of aphlebise and the passage 
of the aphlebia-traces out to these organs (text-fig. 2). The unique opportunity I 
have had of studying numerous petioles of all sizes has resulted in the elucidation of 
the petiolar development. The material on which I have been working is practically 
a felted mass of petrified stems, roots, petioles, pinnse, and pinnules of the species. All 
of them are fragmentary, but, by cutting series of sections from various individuals, 
a more or less continuous chain of development has been established. The extent of 
the overlapping in these series is shown in the following table : — 



Ser/es 


Stage of 

PI / Jiff 6. 


Staffe of 
Pt mftff 33 


Stage of 
PlM.fip.32 


Stage of 
Pimfiff.2$. 


■ \'/,ac/e of 
PlW.fi0.27. 


Stage of 
PlJI.fiff.25. 


1198-1205 




















1249 - 1252 


















1253- 1259 


















1230- 1242 




















1 188 - 1 193 


















1188- 1190 


















1177- 1186 




















1267- 1276 





















TEXT-FlG, 1. — Diagram illustrating the extent to which tlie various series of sections overlap. These series consist 
of sections which only show early stages of the petiole-trace. 

* Phil. Trans. Roy. Soc, vol. clxiv., 1874. 
t Trans. Roy. Soc. Edit)., vol. xlvii., 1910. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 167 

General Structure. 

In order to obtain a general idea of the organisation of this fern — Metaclepsydropsis 
duplex — it will be most convenient to follow its tissues from above downwards. By 
proceeding in this way we shall find that we pass from what is already known to what 
is unknown. In the first place we shall concentrate our attention on the xylem tissue, 
partly because it is better preserved and more continuous than any other, but chiefly 
because it changes very considerably as we descend. 

In the smallest divisions of the pinnae with which we are acquainted — namely, 
tertiary pinnae — the trace is curved in form, with tapering, incurved ends. In the 
sinus, formed between the hooked ends and the body of the trace, the protoxylem 
elements may be found. These bundles then are C-shaped or horseshoe-shaped, and 
they are emitted, alternately on each side, from the ends of a similarly shaped secondary 
pinna-trace. At the point of emission of the tertiary pinna-trace the combined 
trace of tertiary and secondary pinna has four protoxylem groups. Lower down the 
two inner groups die out, and the resulting trace is again C-shaped, with hooked ends, 
and two protoxylem groups in the sinuses formed by these hooks. 

The smallest traces known are distinctly curved, and have parenchyma in the 
concavities. In one or two examples, however, the secondary pinna-trace is not open, 
although it gives off open tertiary traces. In this case the emission of the small strand 
strongly resembles the departure of the secondary pinna-trace from the primary one in 
the genus Clepsydropsis. PI. IV. fig. 42 is a good example of this emission. 

The occurrence of a similar "closed" bundle has also been noted in several primary 
pinnae (PI. IV. fig. 43). This probably only takes place some way along the pinna, 
because, as such primary pinnae are followed downwards into the petiole, each pinna- 
trace becomes like a much flattened C with hooked ends. No primary pinna-trace has 
been observed which was " closed " while still passing through the cortex of the petiole. 
Apart from these few exceptional examples, the pinna-traces of one order join those of 
a lower order alternately on opposite sides as described above. 

When we come to the primary pinnae, however, a change takes place. They do not 
enter the petiole separately but in pairs (PI. IV. fig. 40, pin. tr.), and these pinna-pairs 
enter alternately on each side of the petiole. Two primary pinna-traces, then, enter 
the cortex of the petiole at the same level and pass downwards to join the petiole trace. 
(Before joining the petiole-trace, however, each pair of pinna-traces unite to form one 
xylem arc, as will be explained later (PI. II. fig. 19, b).) They are placed symmetri- 
cally one on each side of the principal plane of the petiole, and are thus " mirror images " 
in this plane of symmetry. A short distance below the point where the two primary 
pinna-traces enter the cortex of the petiole, two pairs of very small traces may be seen 
to pass into the petiole. These are the aphlebia-traces. These also are "mirror 
images " of one another in the principal plane of the petiole. They are situated out- 
side the pinna-traces. A transverse section through a petiole at this level shows seven 



168 



W. T. GORDON 














Text-Fig. 2. — Metaelepsydropsis duplex. Series of transverse sections of petiole showing pinna departure and position 
ofaphlebise. Natural size. After sections 1310-1318, Kidston Collection. 



Z 2 °\ 








Text-Fig. 3. — Formation and gradual reduction of islands of parenchyma in the petiole-trace. The series illustrates : («) the 
formation of the island on the upper ends of the trace, and (b) the formation of the groove on the lower ends. The groove 
is seen dying out on the inner side of the island at the top of the traces. Complete series shown by figures in following 
order: 1, 2, 3, 4, 5, 6 (upper ends), 3, 4, 5, 6 (lower ends), 1, 2, 3, 4, 5, 6 (upper ends, inner side of island). 







1. 2. 3. 4. 

Tkxt-Fig. 4.— Diagram to illustrate distinction between " arms'' and mere dilatations of the arms or of the ends of the trace- 
(1) Dineuron eUipticam, x 45. {2) Metaelepsydropsis duplex, x 12. (3) Diplolabis rumeri,x 7. (4) Etapteris Seoiti,x\\. 
A = arm ; D = dilatation. Slightly diagrammatic drawings from actual specimens. Magnification indicated after name. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 169 

traces — the petiole-trace, two primary pinna-traces, and two pairs of aphlebia-traces 
(text-fig. 2, No. 3, 4, and 5, and PI. IV. fig. 40). 

Lower down in the petiole the four aphlebia-traces unite to form two, or, if we 
momentarily change our point of view and follow the aphlebia-traces upwards, the 
single strands each bifurcate (PL IV. fig. 40, aph. tr.). In PL IV. fig. 39 a section 
at a still lower level is shown. The aphlebia-traces have almost joined on to the outer 
ends of the pinna-traces. In PL III. fig. 38 the junction is complete, but the four 
protoxylem groups on each combined-trace indicate that there has been a fusion of two 
distinct strands. The two inner protoxylem groups die out downwards, and at the 
level of PL III. fig. 37 each combined-trace has two protoxylem groups, one at each end. 

Proceeding still lower, the two combined-traces unite to form the pinna-trace-bar, on 
which there are four protoxylem groups. A short distance below this the bar joins on 
to the hourglass-shaped xylem of the petiole. In PL II. figs. 19 and 18 the bar (b) is 
shown immediately before and immediately after its union with the petiole strand. 
(These figures are not, however, from the same specimen.) An island of parenchyma 
thus appears at the end of the petiole-trace (PL II. fig. 18, is), and round the periphery 
of the island there are six protoxylem groups (PL II. fig. 21). Four of these belong to 
the pinna-trace-bar and two to the petiole-trace. These latter lie on the sides of a 
small bay situated at the end of the hourglass-shaped petiole strand (PL II. fig. 21, 
prx\, prx c z). Still continuing downwards, the island gradually becomes smaller and the 
protoxylem groups of the pinna-trace-bar unite in pairs to form two groups (PL II. fig. 
20, prx u prx. 2 ). The small bay with its two protoxylem groups also gradually dis- 
appears, and the island assumes an elliptical form with a protoxylem group at each apex 
(PL II. fig. 21, prx\, prx\). 

As we descend still further, the island becomes circular ; then it opens to the ex- 
terior again, and an open bay with two protoxylem groups results (PL II. figs. 17, 18, 
19, 22, 23, and 24, g). This bay is vertically below the one which was mentioned 
above, and so the mistake has arisen that there was a permanent groove running down 
the petiole-trace on each side. As has been shown above, the groove ultimately dies out 
and is re-formed lower down ; indeed, it is the last vestige of the wedge of parenchy- 
matous tissue shut in between the entering pinna-trace-bar and the petiole-trace. The 
disappearance of the groove and a diagrammatic representation of the reduction of the 
island are represented in text-fig. 3. 

So far we have seen that two primary pinna- traces and their "spreads"* of pinnae 
of a higher order have joined the petiole-trace and gradually disappeared into it. A 
similar sequence is shown at the opposite end of the petiole-trace, but the two ends are 
never in the same phase of pinna-trace emission. Thus four orthostichies of pinna 
" spreads " are borne by each petiole. Up to this point every stage has been ascertained 
by reference to numerous series. 

* A pinna spread is the whole assemblage of secondary pinnte^borne by a primary Jpinna and spread out into 
one plane. 



170 W. T. GORDON 

In following out the changes in the petiole-trace much greater difficulty has been 
met with, and the stages figured are not from long series. This was due to the fact that 
the material consisted of short lengths of petioles, no specimen exceeding 6 inches. In 
that distance the changes were very slight, so that long — very long — series would pro- 
bably be necessary to show the gradual change from one of the stages figured, to the 
next. In spite of this discontinuity, the position I take up will be readily accepted, for 
there is sufficient evidence to prove the general problem, namely, that there is a gradual 
disappearance of the "waist"* or constriction in the middle of the hourglass-shaped 
petiole-trace, and that this reduction takes place as we descend in the petiole. The 
last part of the problem is much more difficult to prove than the first, but, concurrently 
with the disappearance of the " waist" of the petiole-trace, there is a marked reduction 
in the size of the whole trace and in the pinna-traces which join it. We know that the 
petiole-trace in other ferns gradually becomes smaller as we descend towards the stem, 
and that there are reduced pinna-traces towards the base of the petiole. This is what 
one would naturally expect. It is not surprising, therefore, to find a similar condition in 
Metaclepsydropsis. But apart from the xylem tissue, we find that the outer cortex 
loses its sclerotic outer layer as we follow this series. Now, from a study of the stem 
cortex, which has no sclerotic elements, we should naturally expect that the free petiole 
near its base would have no sclerenchyma in the outer cortex. From this evidence also 
we must conclude that the series, as figured, shows the various stages in a descending 
order. In other words, the proof of the problem lies in the following observations : — 
The size and shape of the petiole-trace, the development of the pinna-trace-bar and 
pinna-traces, and the sclerotic outer cortex are all gradually reduced as we proceed 
through the figures shown from PI. II. fig. 17 to PI. III. fig. 34. 

It might be said that this was not conclusive, since all the figures do not represent 
different levels in one and the same specimen, and that there are perhaps several 
different species mixed up in the series. Such an objection can easily be met, for each 
stage figured is intermediate between its two neighbouring figures ; there is no sudden 
jump at any point from one type of trace to a totally different one. In fact, there is a 
continuous variation in one direction, and this is the only variation that is shown by the 
specimens. 

PI. II. fig. 17 represents a petiole-trace perfectly typical of the species Metaclepsy- 
dropsis duplex ; the waist is very pronounced, and the pinna-trace-bar is also well 
marked. The specimen shown in fig. 18 is also perfectly typical though rather smaller, 
but, apart from a few details in the phase of the pinna-trace departure, the two figures 
represent specimens which are specifically identical. In fig. 19, however, the waist is 
much less marked than in figs. 17 and 18, while fig. 20 exhibits a trace with hardly 
any constriction in the middle. There is still a slightly sinuous appearance in fig. 21, 
but in figs. 22 and 23 this has entirely disappeared. Indeed, had the petiole whose 
trace is shown in the last-mentioned figure been discovered separately, there is no doubt 

* The word " waist " is used here as it gives the idea of a constriction in the middle of the trace. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 171 

that it would have been referred to another genus, namely, Dineuron. Fig. 24 repre- 
sents another example where the trace is even more like Dineuron than that of fig. 23. 

The next two figures (25 and 26) are cut rather obliquely, and the traces are perhaps 
not really qaite so long as they appear, but the angle of section is very near 90°, so that 
any reduction in their size on account of the obliquity of section must be very slight. 
They are from different specimens, but represent stages which are practically identical 
and show the departure of a much reduced pinna-trace-bar (b). Fig. 26 is probably 
further up the petiole than fig. 25, and the bar (b) shows clearly that it will divide into 
two portions. A somewhat similar example, cut from the same specimen as fig. 26, is 
shown in PI. III. fig. 27, though here the trace is rendered shorter than it really should 
be by a slight crushing. In PL III. fig. 28 the trace is still further reduced, but it is 
clearly identical with that of fig. 27, while that shown in fig. 29 is also similar. 

PI. III. figs. 30 and 31 represent examples quite comparable with that shown in 
fig. 29. 

The next three figures represent sections taken from the same specimen, and the 
petiole joins the stem two sections below that shown in fig. 34. Fig. 32 is similar to 
fig. 31, except that both ends of the trace are closed instead of only one end, as in the 
former figure in which the left-hand end exhibits an island of parenchyma in the xylem, 
while the right hand end has a bay. Fig. 33 shows clearly the two almost circular 
islands, one at each end of the trace, and also the stem (st) which the petiole joins at 
a lower level. Near the junction with the stem the islands disappear and only the 
small protoxylem elements remain. These protoxylem groups occur in pairs at each 
end in fig. 33, but in fig. 34, where unfortunately the lower part of the trace is crushed, 
there only appears to be one group. Two sections lower down in this series the petiole- 
trace joined the stem (figs. 33 and 34, st). The petiole-trace of M. duplex has thus 
been followed through all its stages into a stem which, while presenting certain 
peculiarities, is very simple in structure. 

A typical transverse section of this stem is shown in PI. I. fig. 1 or PL IV. fig. 45. 
Two regions may at once be distinguished, an outer of large tracheides which are seen 
to be reticulately thickened, when viewed in longitudinal section, and a central zone 
composed of a mixture of tracheides and parenchyma. The inner tracheides are long, 
pointed elements and have reticulate or scalariform thickenings on their walls, but 
they are smaller in diameter as a rule than those of the outer zone. The stem was 
long and dichotomously branched. 

On the whole it had the appearance of a rhizome, and evidence in favour of this will 
be brought forward immediately. 

Root-traces have been met with on one or two occasions, but only one example 
actually joined the stem xylem. One of these root-traces is shown in PL IV. fig. 44. 
It is large for a root-trace and is diarch. 

So far merely the xylem tissue has been considered, but the cortex is also of some 
importance ; for, while there is a sclerotic layer in the cortex of the mature petiole, there 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 8). 27 



172 W. T. GORDON 

is none in that of the stem or of the lower parts of the petiole. The inner cortex has 
generally all decayed, but small patches sometimes appear in the islands formed between 
the pinna-trace-bar and the petiole-trace. It has always the appearance of a delicate 
parenchyma. 

Numerous groups of sporangia, probably synangia, occur scattered through the 
blocks containing the petioles and stems of M. duplex. There are generally four 
sporangia in each group, but there is no evidence that these were borne by the pinnules 
of this plant 

Histology of the Stem. 

The stem which has just been referred to M. duplex is of peculiar interest both on 
account of its simple structure and because it is of considerable phylogenetic import- 
ance. In transverse section it is circular in outline (PL I. fig. 1, and PL IV. fig. 45), 
and about 1*8 mm. in diameter. Two zones may easily be distinguished, the inner 
consisting of a mixture of parenchyma and tracheidal tissue, and the outer being made 
up entirely of tracheides. PL IV. fig. 45 probably gives the best idea of the distribu- 
tion of the tracheides and parenchyma in transverse section. (Although a petiole-trace 
connected to the stem has been cut away at the bottom of the figure, it has not disturbed 
the more central tissues.) In other examples the tracheides are almost absent, and in 
PL IV. fig. 47 such a specimen is figured ; only one tracheide (t) is present, and it is 
immersed in the centre of a thin-walled parenchyma (p). The general circular shape of 
the stem is distorted in this specimen, and a petiole-trace is shown connected with the 
stem. This trace apparently belonged to a petiole which had been torn away, leaving 
a ragged stump still connected to the stem, and the tissues of the trace are so crushed 
that one of the protoxylem groups is unrecognisable, and the whole trace very irregular 
in outline. 

In most specimens, however, the parenchyma has been ruptured and only fragments 
of the delicate cell-walls remain. As a result the tissue is seen better in transverse 
than in longitudinal section. In one example (PL IV. fig. 46) a peculiar condition was 
discovered. There is no parenchyma present in the stem, and there is a distinct radial 
arrangement of the tracheides at a. This stem is very much smaller than any of the 
others, and I have only seen it in one preparation. 

I am inclined to think that it is an unequal dichotomy of the stem, and it will be 
referred to later. In the meantime the absence of conjunctive parenchyma is worthy 
of notice. 

The tracheides of this inner zone are generally small in diameter and vary consider- 
ably. The largest are only between 45m and 50/x, while the small ones are 30m and 
under. Those nearest the periphery are generally the largest, and they abut on the 
inner tracheides of the outer xylem zone, which are not much larger than the outer 
tracheides of the mixed pith. The thickenings on the walls of the inner elements are 
scalariform in the smallest but reticulate in the largest. In the same tracheide the 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 173 

transition between scalariform and reticulate thickening is often clearly seen (PI. I. fig. 2). 
The largest of these elements have rarely more than three rows of pores on each wall. 

The outer zone of the stem xylem is four to five cells deep and constitutes a ring of 
solid wood round the inner zone. The elements of this ring are much larger than those 
of the inner wood, but the tracheides towards the inner margin of the ring are smaller 
than the others as a rule. The average size is from 100m to 130m, but near the bifur- 
cation, to be noted shortly, some of the elements are as much as 250m in diameter. 
Some of these large tracheides may be seen in PI. I. figs. 4, 5, and 6. At the departure 
of a petiole-trace larger elements also occur (PI. I. fig. 8). Like those of the inner zone, 
the outer tracheides are long, pointed, and their walls reticulately thickened, but there 
are six to nine rows of pores on the walls. The pores on the walls of the elements of both 
zones are about the same size. Groups of small tracheides may occasionally be seen in 
the outer xylem zone, but these are the decurrent protoxylem elements of a petiole-trace 
and do not belong to the stem itself. 

Outside the xylem there is generally a dark layer which in very thin section is seen 
to consist of cellular tissue. This tissue no doubt represents inner cortex and phloem, 
but it seems to have been composed of very delicate elements, which are too much 
crushed to warrant more than a passing notice. The outer cortex, on the other hand, 
is generally well preserved. It consists of a thick- walled parenchyma, but there are 
no sclerotic layers present. This outer cortex is sometimes produced into ragged 
protuberances. 

To the stem xylem, petiole- and root-traces are attached, but they only occur at 
considerable distances from each other. In the material examined only three or four 
of such emissions have been noted, and in but one case was a root-trace found joining 
the stem. In the case of each of these appendages, however, there is sufficient evidence 
to show that the departure is protostelic, i.e. the outer xylem ring is never broken, 
thus exposing the inner zone on the surface of the cylinder. 

Branching of the Stem. 

In several of the stems examined it was possible to cut long series of sections, but 
the changes were very slight, except in the neighbourhood of the emission of a petiole- 
or root-trace. In one specimen, however, just above the point of departure of a petiole- 
trace, I was fortunate enough to discover both a root-trace emission and a bifurcation 
of the stem, within a length of l|- inches. In another case, just above a petiole-trace 
departure the very small solid stele of PI. IV. fig. 46 was discovered, and this is possibly 
an unequal dichotomy of the stem. 

PL I. fig. 3 represents the first stage of such a bifurcation (the small trace rt. tr. 
represents a root-trace which arches over at this point, and the top of the arch is here 
shown), and the inner zone of mixed pith is distinctly elongated, as is the whole stem 
xylem. Roughly speaking, the stem xylem is an ellipse. Higher up the stem the 



174 W. T. GORDON 

mixed pith divides into two large masses connected by a thin neck, or, in other words, 
becomes figure-of-eight-shaped, the outer xylem still remaining elliptical. As a result 
of this, the large elements previously noted are found towards the outside of the narrow 
neck connecting the two masses of the mixed pith (PI. I. figs. 4 and 5). The pith then 
divides into two separate masses and the outer xylem becomes figure-of-eight-shaped, 
with large xylem elements in the constriction of the eight (PI. I. fig. 6) ; and finally 
the two parts separate. The division is into two equal parts, i.e. it is an equal 
dichotomy. During the division there is no appearance of a branch gap, i.e. the 
departure of the branch is protostelic. Thus all emissions from the stern, whether of 
petiole, root, or branch-trace, have proved to be protostelic. The small solid branch — 
if it be a branch — is of interest because it is quite similar to a case noted in Diplolabis 
romeri (Solms), where the inner wood almost entirely disappeared at a bifurcation of 
the stem. * Such a reduced branch also might be expected to show primitive character- 
istics and to give some indication of the race from which the plant had sprung. The 
evidence from this specimen points to an ancestor with a solid stele, and this is in 
harmony with the evidence from a study of the petiole. 

Histology of the Petiole. 

It is generally admitted that, at the junction of the petiole with the stem, ancestral 
characters may be expected, so it is very important that the changes in the petiole-trace 
near the base should be carefully noted. In making a minute examination of the xylem 
tissue in this region, it will be found more convenient to work up the petiole, and not 
down as in the general description. 

The first figure to be noted is PI. II. fig. 16, which represents a transverse section 
of the stem in a rather flattened condition. At a a short arm of parenchymatous tissue 
may be seen stretching from the central mixed pith into the outer xylem. Near the 
end of this radial arm are some rather small elements which constitute one of the 
protoxylem groups of a petiole-trace. The other protoxylem group is not differentiated 
until later, so that, at this early stage, the difference in phase of the two ends of the 
petiole-trace is quite marked. In the next section of the series (PI. II. fig. 15) the 
arm a is much longer, and the protoxylem group (prx 1 ) is at the end of the arm furthest 
from the inner zone of mixed tracheides and parenchyma. There is still no sign of 
the second protoxylem group. Fig. 1 4, which follows, shows one group of protoxylem 
elements isolated in the outer xylem zone, the arm of parenchyma and its accompanying 
small tracheides having disappeared. On the inner margin of the outer wood, a short 
distance round from where the first medullary arm appeared, a second small sinus 
containing protoxylem may be seen at a'. The next section, represented by fig. 13, 
shows the first protoxylem group (prx 1 ) well out in the outer zone and the second 
group {prx 2 ) quite distinct. 

In PI. I. fig. 1 2 both groups are clearly seen, and in the next two sections (figs. 1 1 

* Gordon, "On Diplolabis romeri (Solms)," Trans. Roy. Soc. fidiu., vol. xlvii. p. 720, 1911. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 175 

and 10) little change can be noticed. One group has come to lie close to the margin 
of the stem xylem, while the other is at the outer extremity of the arm of parenchyma 
and small tracheides. When the level of fig. 9 is reached, however, the second arm 
has disappeared, and at that end of the stem xylem there are two sunk protoxylem 
groups. At the same time the whole end of the stem has a distinct bulge on it. One 
section higher up (fig. 8) the bulge on the stem is still more pronounced ; indeed, the 
petiole-trace has now become a distinct body connected to the stem at J by a number 
of very large tracheides. The protoxylem groups are placed at opposite foci of the 
elliptical trace. 

In the section above that shown in fig. 8 the trace has separated from the stem 
as an elliptical mass (fig. 7). This trace was followed outwards for three sections, but 
no change was noted, and above that level it had disappeared. The subsequent changes 
in the petiole-trace will be followed in another series which starts with a much-crushed 
section just about the level shown in PL I. fig. 8. Below this level both stem and 
petiole have rotted away. The first section of this new series which is figured, is shown 
in PI. III. fig. 34. Here the much-distorted petiole-trace is closely adpressed to the 
stem xylem, and only one protoxylem group can be seen. This group is quite similar 
in position and shape to either of those in PI. I. fig. 7. A few sections higher up, 
however, each protoxylem group has divided into two (PI. III. fig. 33, prx), and an 
island of parenchyma is developed at each end, separating the two parts into which 
each protoxylem group has divided (fig. 33, is). Fig. 32 represents the highest section 
in this same series, and at the top of the trace there is a mere filament of xylem tissue 
on the outer margin of the island of parenchyma. A closely similar specimen is 
represented by PL III. fig. 31, where one end of the trace is open while the filament 
is just leaving the other end. There has been some slight lateral crushing in this 
specimen, as is indicated by the flattened cortex, but it has not affected the trace 
very much. 

In PL III. fig. 29, however, a similar trace is shown, but in this case there 
appears to have been no distortion. The left-hand end of the trace is open, and shows 
the bay g with the protoxylem elements at each side. The other end is closed and, 
there also, the two groups of protoxylem are clearly seen. So far no distinct pinna- 
trace-bar has been detached from the petiole-trace ; though the filaments no doubt 
represent very reduced examples of such departures. The trace shown in PL III. 
fig. 28, while practically identical in form with that of fig. 29, is slightly larger, and on 
the left-hand end the filament of xylem of fig. 29 is represented by a stouter bar. 
Unfortunately these petioles were only short fragments, generally less than half an inch 
long, so that the pinna-trace-bars could not be followed very far. In one example, 
however, from which about twenty sections were prepared, two of these bars were 
followed a short distance, but they never, as far as could be seen, passed out into distinct 
pinnae. Though these figures are taken from different specimens, it will be noticed that 
as the trace increases in size the pinna-trace-bar becomes more and more robust. 



176 W. T. GORDON 

PI. III. fig. 27 represents a cross-section of another of these petiole fragments which 
is exactly similar to that of fig. 28. The one end is open and the other closed by a 
pinna-trace-bar in exactly the same condition as in the latter figure. At a higher level 
in the specimen the pinna-trace-bar became detached from the petiole- trace and showed 
some signs of dividing into two (PI. II. fig. 26, b). A twisting of the whole petiole- 
trace at this level gives a much longer appearance than is shown in PL III. fig. 27. In 
the next figure (fig. 25) an exactly similar section is represented. The plane of section 
in this case is horizontal, except at the left-hand side. In this specimen there is also a 
reduced pinna-trace- bar given off from the petiole-trace, and, as in the case of fig. 26, the 
bar shows signs of dividing in the middle to form two traces. 

Passing to the trace represented in PL II. fig. 24, the first thing to be noticed is 
that it has suffered no distortion and that it is essentially similar to that of fig. 25. In 
general form it greatly resembles PL III. fig. 29, but it is larger and the pinna-trace-bar 
b at the top of the figure is much better developed than in the former specimen. The 
petiole from which the section was prepared is about 6 inches long, and a little above 
the level of this section two pinna-traces were noted. They were fairly well developed, 
but did not penetrate the cortex to enter into distinct pinnae. The outer cortex (o.c.) 
in the figure is also worthy of notice, since it contains no sclerotic layer. 

A very similar example is shown in PL II. fig. 23. The trace is not quite so long 
as the last, but it is much stouter, and the bar b is rather better developed. Yet this bar 
also never gives rise to traces which enter into distinct pinnae. In fig. 23 it is practically 
divided into two, but both parts die out higher up. The outer cortex in this specimen 
had a distinct sclerotic band towards the periphery. Up to this point all the traces 
examined would have been referred to Dineuron had they occurred separately, and 
indeed would still be referred to that genus unless the next three stages had been 
discovered. The first of these transition forms is shown in PL II. fig. 22. It is 
essentially like that of fig. 23, but a flattening is making its appearance in both sides 
and the pinna-trace-bar is rather well developed. A similar flattening may be noted in 
PL II. fig. 26, but this is largely due to the obliquity of the plane of section, which causes 
an apparent elongation of the trace. PL II. fig. 21 exhibits another of these transition 
stages. Here the trace gives some indication of a waist. The pinna-trace-bar, also, is 
much more robust than in the last figure. This trace is exactly intermediate, both 
in size and shape, between the Dineuron-like example of fig. 24 and the normal 
Metaclepsydropsis trace of figs. 17 and 18. 

In the specimen shown in fig. 20 we have the last of the transition types. A 
distinct waist is shown in this example, and therefore the trace, in transverse section, 
has the appearance of an hour-glass. An early stage of pinna-trace departure is also 
demonstrated by the specimen, and the bar is quite robust and well developed. Probably 
even this specimen would only be referred to Metaclepsydropsis duplex with some 
misgivings. Fig. 19, however, supplies the last link in the chain, and, while the trace 
shown in it would at once be accepted as typical of Metaclepsydropsis duplex, there 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 177 

is one peculiarity, the broad waist, which links up the trace of fig. 20 with those of 
figs. 17 and 18. 

While it would be absolutely impossible to pass from the trace of PI. II. fig. 17 to 
that of PI. I. fig. 7, or even that of PI. II. fig. 24, yet when all the intermediate stages 
are shown it becomes not only possible but quite simple. The whole development may 
be summed up thus : — The demands made on the xylem of the petiole by more and more 
robust pinna-traces have caused the petiole-trace (l) to become larger, and (2) to increase 
the size of the ends so as to accommodate the much stouter pinna-trace-bar. A closely 
similar conclusion was arrived at in the case of the petiole-trace in Diplolabis romeri, 
but instead of the ends of the trace becoming inflated they were drawn out into two 
long arms at each end. 

In PI. III. fig. 36 an interesting example is shown. The specimen was very short, 
but it happened to include the portion of the petiole to which the primary pinnae are 
attached. These pinnae were seen to be quite normal though small. This has led me 
to believe that the specimen must have been a portion near the top of a petiole. In 
that region we would naturally expect the trace to be smaller, and, if the hypothesis be 
correct that the hour-glass shape was impressed on the petiole-trace in order that it 
might accommodate more robust pinnae, we would also expect that the petiole-trace 
near the apex would revert to its elliptical form, since the pinnae diminish in size near 
the top of the frond. The trace in this example is exactly similar to that in PI. II. 
fig. 21, except that it is smaller, and the latter figure represents a section well down 
towards the base of the petiole. 

Turning now to examine the primary pinna -trace, we find that one of the earliest 
obvious stages in its development on the petiole-trace is seen in PL IV. fig 39, where, at 
g, a round island of parenchyma is seen. At each side of this island there are groups 
of small elements — the protoxylem. Such a stage, however, is not the earliest that may 
be seen, for the protoxylem groups may be observed before the island is formed, i.e. 
when there is a groove at the end of the petiole-trace and not an island (PL II. figs. 
17, 18, 19, 22, 23, 24, prx\, prx a 2 ). (The island has, of course, been formed by the 
production of xylem elements at the horn-like extremities of the sides of the groove. 
When the xy]em outgrowths from the two sides meet in the centre the groove becomes 
an island of parenchyma.) The protoxylem groups may be distinguished at a still 
lower level, where the groove has become very shallow ; indeed, the groove is a mere 
depression between the two protoxylems (PL II. fig. 21, prx\, prx c 2 , and fig. 17, prx\, 
prx b 2 ). PL II. fig. 18 is probably just below the level at which the last vestiges of 
these protoxylem groups are visible, although the bay or groove which results from the 
island opening to the outside is still seen to persist at C. 

A stage beyond that of fig. 39 is shown in PL II. fig. 21, at the lower end of the trace. 
The island of parenchyma which was circular in the former figure is elliptical in the 
latter, and has still two protoxylem groups, one at each end of the major axis of the 
ellipse. At a still higher level (PL IV. fig. 39, loiver end of trace) the island is much 



178 W. T. GORDON 

larger, and the protoxylem groups of the pinna-traces, vertically above the pair re- 
presented by the pinna-trace-bar we are considering, are quite distinct on the inner 
margin of the island. A somewhat similar example may be seen at the top of the trace 
in PI. II. fig. 20. There are at this stage four protoxylem groups arranged round the 
island. The groups at the ends of the island then spread out on the outer margin and 
afterwards divide into two. The stage just before the division is shown in the last- 
mentioned figure. 

Passing to the top of the trace in PI. II. fig. 22, it will be noticed that the pinna- 
trace-bar has four protoxylem groups peculiar to itself. These have been derived as 
indicated above from the two original groups. There are thus six groups round the 
island, which has now attained a larger size. Two closely similar examples are figured 
in PI. II. fig. 21 and fig. 18, at the top of the trace in each case. In figs. 17 and 19 the 
bar has become detached from the petiole-trace and gives some indication that it will 
divide into two equal parts. PI. III. fig. 37 represents a still later stage where the 
division is complete and two curved xylem strands are produced. In the next section, 
other protoxylem groups make their appearance on the inside of each of the curved 
xylem strands, but they are near the lower ends of these strands and not in the centre. 
Two sections above this last one these extra protoxylems are distinct, and the arc of 
xylem is converted into a double arc (PI. III. fig. 38,pi7i. tr., aph. tr.). In PL IV. fig. 39 
— about three sections above fig. 38 — the small traces are cut off from the larger median 
bundles. The small bundles pass out to supply aphlebise, and during their passage 
outwards they each bifurcate. The two branches of each bifurcation pass out at the 
same level (PL IV. fig. 40, aph. tr.*, aph. tr. a 2 ), and not, as in Diplolabis romeri, at 
different levels. 

The larger median bundles — destined to supply pinnae — become more curved (PL IV. 
fig. 41), and finally pass out at the same level into the pinnae, of which there are two for 
each pinna-trace-bar. In some cases, however, the incurved ends unite, and the arc then 
becomes a closed ring of xylem (PL IV. fig. 43). Such a closed trace is very interesting 
and probably indicates an ancestral character. Very few of such annular traces have 
been discovered, and unfortunately they could not be followed far enough to see if 
they ultimately resumed the open form, though they probably do resume the horse- 
shoe shape higher up. In one secondary pinna-trace a similar character was noted, 
and here the emission of the tertiary pinna-trace is exactly comparable with the 
emission of the secondary pinna-trace from the primary in Clepsydropsis antiqua, as 
shown by Dr P. Bertrand.* 

Apart from these abnormal cases the pinna-trace is open and consists of reticulately 
thickened tracheides with scalariform protoxylem elements in the sinuses formed by 
the incurved ends. The pinnae of one order are cut off from the ends of the trace 
of lower order, and pass out alternately at each side. 

The foliage and fructifications of this species are quite unknown, though several 

* P. Bertiiand, Etudes but la fronde des Zygnjite'ride'es, 1900. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 179 

groups of sporangia occur in the same sections as contain the petioles and pinnae 
of the plant. In the absence of evidence of continuity, however, it would not be 
safe to connect the two. 



Histology of the Root. 

The root-trace is very similar to that of Diplolabis romeri (Solms) ; it is barrel- 
shaped and consists of reticulately thickened tracheides. There are two protoxylem 
groups of scalariform elements situated one at each end of the barrel-shaped trace. 
In one specimen a rootlet seems to be given off from the main trace, but it is very 
much crushed, and one cannot determine whether the smaller rootlets were similar to 
the large ones. Another specimen curved upwards at first, but ultimately turned and 
grew in the opposite direction to that in which a petiole-trace was emitted. Very few 
root-traces were discovered, and of these only the xylem was preserved. 



Comparison with other Species. 

In their memoir on the fossil Osmundacese, Kidston and Gwynne- Vaughan have 
divided the Zygopteridese into three great groups, and, although few stems belonging 
to this family are known, at least one of the recorded specimens belongs to each group. 
Ankyropteris corrugata (Williamson), A. Brongniarti (Renault), and probably A. 
sca?idens (Stenzel) belong to the first group ; Diplolabis romeri (Solms) and Meta- 
clepsydropsis duplex (Williamson) to the second ; and Etapteris di-upsilon 
(Zygopteris Grayi) (Williamson) to the third. The stele of the stem in all of these 
species is either circular or roughly stellate in transverse section. When examined 
in detail, however, there are considerable differences between them. 

In comparing the newly discovered stem of Metaclepsydropsis duplex with the 
others, we shall begin with the simplest known type (it is also the oldest known type). 
Diprtolabis romeri. This latter species has been recorded from the same locality as 
M. duplex, viz. Pettycur, as well as from other localities in France and Germany 
where M. duplex does not occur. In it the stem xylem is circular in transverse 
section and consists of two kinds of tracheides, both of which have reticulate thickenings 
on their walls. The xylem elements are arranged in two zones, the inner of which 
contains only short, square-ended tracheides, while the elements of the outer zone 
are long and pointed. The inner tracheides are smaller in diameter than those of the 
outer zone, and they are arranged in vertical series as though they had been derived 
by the septation of long elements. There is no conjunctive parenchyma present in 
the stele. The stem branched dichotomously, and was probably a rhizome. 

The departure of the petiole-trace from the stele is protostelic, and the trace is at 
first elliptical, with sunk protoxylem groups, one near each focus of the ellipse. In the 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 8). 28 



180 W. T. GORDON 

subsequent development of the petiole-trace long arms are gradually produced, and 
the trace then becomes H- or X-shaped. Primary pinnse depart from the petiole 
in pairs, one pair alternately on each side, so that four rows of primary pinnae are 
produced on the one petiole. The root-traces also leave the stem xylem in a 
protostelic manner ; they are diarch and barrel-shaped. 

As we have seen, the stele of M. duplex is also circular in outline, and the xylem 
consists of two kinds of tracheide. Both kinds are long, pointed, and reticulately 
thickened, except the smallest in the inner zone, which have scalariform thickenings 
on their walls. There is conjunctive parenchyma present, however. The stem 
branched dichotomously and was a rhizome, as in the last-mentioned species. 

The petiole-trace leaves the stem-stele in a protostelic manner as an elliptical mass 
with two sunk protoxylem groups, but its subsequent development is quite different 
from that of Diplolabis romeri. No arms are produced, but both ends of the trace 
become dilated. In D. romeri the island of parenchyma enclosed by the entering 
pinna-trace-bar is constant in size, whereas that in M. duplex gradually diminishes 
until it is exceedingly small, when it opens to the exterior, and a small groove is 
left in the petiole-trace instead of the wide V-shaped groove between the arms in 
D. romeri. 

The primary pinnae are borne in four orthostichies just as in Diplolabis, but 
occasionally the trace becomes closed as in Clepsydropsis. The roots of M. duplex 
are also quite similar to those of D. romeri. I have gone into considerable detail in 
this comparison, as these two species have many points of similarity and form two 
important links in a possible chain of evolution among the Zygopteridese. 

In comparing M. duplex with Ankyropteris corrugata, a great similarity in the 
shape of the stele must be noted. Both have two zones of tracheides, and there is con- 
junctive parenchyma in the inner zone in each. But while only a few of the elements 
in the inner zone in M. duplex are scalariform tracheides, all the elements in both zones 
of A. corrugata have that type of thickening. It has also been shown that the arms of 
parenchyma and tracheidal tissue which radiate out from the inner zone of the stem- 
stele do not persist for any great vertical distance in M. duplex. In the case of A. cor- 
rugata, on the other hand, these arms seem to persist for a greater vertical distance — so 
much so, that the outer zone of the stele-xylem appears to consist of five groups of 
tracheides alternating with five parenchymatous arms which project from the inner zone. 
The largest tracheides in A. corrugata appear in the centre of each group, and in 
M. duplex the largest are in the central portion of the outer xylem ; but if the radiating 
arms broke up the outer ring as in A. corrugata the smaller tracheides would be situated 
along the sides of the arms, and then the large tracheides would occupy a central position 
in the resulting groups of the outer xylem zone. In both species these radiating arms 
of parenchyma and tracheides are intimately connected with the petiole-trace departure, 
and this emission is more distinctly protostelic in character in M. duplex than in A. 
corrugata. In the type of branching shown by both species there is a striking simi- 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 181 

larity ; it is a dichotomy in each case, and in neither case has any " axillary " branching- 
been recorded. It is also worthy of notice that the Botrychioxylon-like specimen shown 
in PI. IV. fig. 46 is paralleled in one of Williamson's figures of A. corrugata* 

The petiolar development, however, is entirely different in the two species, A. cor- 
rugata having only two rows of primary pinnae, and these, as Kidston and Gwynne- 
Vaughan point out,+ being little more than scale leaves. 

The French Permian species A. Brongniarti is also very similar to M. duplex as far 
as its stem-stele is concerned. In size they are almost identical, but in A. Brongniarti 
the radiating arms are exceedingly well developed and the tracheides have all scalariform 
thickenings on their walls. Very little is known about this species, particularly about 
the petiole-trace, but the petiole referred by Renault to the species is distinctly of the 
Ankyropteris type. The branching of the stem is of the " axillary " type now shown to 
be of the nature of an unequal dichotomy. 

Compared with Etapteris di-upsilon [Zygopteris Grayi), the stem of M. duplex is 
very distinct. The stellate structure of the stem xylem in the former species is the 
result of the rapid emission of petiole-traces, and, when one has departed, the outline 
becomes rounded and not stellate at that part of the stem. The arms of parenchyma 
and tracheides radiating from the central zone of the stem xylem are much more 
prominent here than in A. corrugata or A. Brongniarti; and the type of branching is 
an unequal dichotomy. The petiole-trace which is emitted from the stem belongs to 
the third group, as defined by Kidston and Gwynne-Vaughan. This correlation of the 
petiole known as Etapteris di-upsilon (Williamson) with Zygopteris Grayi has only 
recently been published by Dr Kidston. j The species is therefore distinct from 
M. duplex in all its salient features. 

Ankyropteris [Zygopteris scandens) (Stenzel) is another Zygopterid fern with a 
stellate stem-stele. It was in this species that the presence of the " axillary " branch 
was first demonstrated, and in some respects it is closely similar to Etapteris di-upsilon. 
The petiole-trace, however, is of the Ankyropteris type. It is thus quite distinct from 
M. duplex. 

Quite apart, then, from the appearance of the petiole-trace, M. duplex may be dis- 
tinguished from all other species of the Zygopteridese by the structure of the stem. As 
regards the petiole-trace itself, there is no species so far described with which it is likely 
to be confounded in its mature stages ; but, as has already been shown, the early stages of 
the petiole might quite well be confused with Dineuron. It is true that no specimen 
of Dineuron has been discovered which has as large a stele as M. duplex, but it is con- 
ceivable that some species of Dineuron may be discovered with a petiole-trace as large 
as that of the early stages in the trace of M. duplex. None of the early stages of 
petiolar development in the latter species is comparable with Clepsydropsis. In 

* Williamson, Phil. Trans. Roy. Soc, vol. clxvii., 1876, pi. v. fig. 19. 

t Kidston and Gwynne-Vaughan, Trans. Roy. Soc. Edin., vol. xlvii., 1910. 

X Ann. Bot., vol. xxiv., April 1910. 



182 W. T. GORDON 

Diplolabis, on the other hand, there was a distinct resemblance to that species (Clepsy- 
dropsis) in the early form of the petiole- trace. 

Summary. 

In this paper a fern stem is described, and the evidence for referring it to Meta- 
clepsydvopsis duplex (Williamson) is cited. The stem is shown to be a long, 
dichotomously branched rhizome, from which petioles and roots are only emitted at 
considerable intervals ; indeed, although several inches of one stem were cut into 
transverse sections, only one petiole departure was observed, and Miss Benson followed 
another stem for nearly 2 feet without discovering any such emission. Several pieces 
of the stem were discovered. Branching of the stem is shown to be of the nature of 
a dichotomy. 

The petiole-trace is followed from the earliest stages of its differentiation from the 
stem until it departs as a separate trace. From that level it is followed through several 
series until the normal form of the M. duplex petiole-trace is attained. The Dineuron- 
like type of the early petiole-trace is recorded. 

The pinna-trace departure is next studied, and the changes noted from the time 
the pinna-trace-bar leaves the petiole-trace until the tertiary pinnae are reached. 
Several cases of abnormal traces are noted in which the xylem forms a closed ring. The 
resemblance which such an abnormal type has to what is the normal form in Clepsy- 
dropsis is commented on. 

Koot-traces are also observed, and they are diarch, with a barrel-shaped xylem mass. 

The cortex is dealt with in general terms, as there is nothing of special note in its 
organisation. In the cortex of the stem and early stages of the petiole, however, there 
is no sclerenchyma present. 

One specimen of the stem is recorded which shows secondary thickening, but it is 
probably abnormal. It appears similar to Botrychioxylon. 

The species is then compared with the members of the Zygopteridese which 
resemble it. 

Metaclepsydropsis duplex (Williamson). 

1874. Rachiopteris duplex, Williamson, Phil. Trans. Roy. Soe., vol. clxiv. 

1889. Asterochlmna (Clepsydropsis) duplex, Stenzel, Die Gattung Tubicaulis. 

1896. Clepsydropsis sp., Renault, Bass, hoaill. et perm. d'Aulun et d'lZpinac. 

1909. Metaclepsydropsis duplex, P. Bertrand, A'tudes sur la fronde des Zygopteridees. 

1910. ,, ,, Seward, Fossil Plants, vol. ii. 

Diagnosis. 

Stem long, dichotomously branched ; xylem of stem circular in transverse section ; 
2 mm. in diameter ; elements in two zones. Inner zone of long, narrow, reticulately 
or scalariformly thickened tracheides, together with some conjunctive parenchyma. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 183 

Arms of inner zone radiate into outer zone in certain places. Outer xylem of long, 
broad, reticulate tracheides. Protoxylem groups, of scalariform elements, not cauline 
but decurrent from the petiole into the stem, disappear at outer margin of inner xylem 
zone. Cortex without sclerenchyma. 

Petiole-trace at first elliptical in transverse section, with a sunk protoxylem group 
near each end. Ends ultimately enlarged and trace becomes hourglass-shaped, with 
sinus (groove) at the middle of each end. Protoxylem groups, two at each end, situated 
on sides of sinus. Tracheides with reticulate thickenings, except protoxylem elements, 
which are scalariform. Primary pinnae in four orthostichies ; inserted on petiole in 
pairs. Tertiary pinna-traces join secondary pinna-traces in two orthostichies. 
Insertion of secondary pinnae on primary similar to tertiary on secondary. Trace of 
primary pinna sometimes closed, but open lower down. Last traces to join primary 
pinna-traces are from aphlebiae. Aphlebia- traces bifurcate during passage through 
petiole-cortex, and both branches pass out at the same level. 

Pinna- and aphlebia-traces unite to form pinna-trace-bar, which joins petiole-trace. 

Root-traces large, barrel-shaped, diarch. 

Foliage and sporangia unknown. 

Localities. — Calciferous Sandstone Series ( = Culm), Pettycur, Fife ; Culm of 
R^gny.* 

Conclusions and General Considerations. 

The discovery of stems of Zygopterid and Osmundaceous affinity has recently 
thrown considerable light on the stelar evolution in both these families. As a result 
of research along this line, the occurrence of a comparatively simple stem-stele in M. 
duplex does not come as a surprise. In only one case has a simpler stele been recorded, 
viz. in Diplolabis ro'meri. The elements of the stem xylem in M. duplex also show 
an archaic type of thickening on their walls — a reticulate type — and in this they are 
similar to the tracheides in the stem of Diplolabis romeri, while quite distinct from 
the xylem elements in the xylem of other Zygopterid stems. In the inner zone of the 
xylem, however, some of the elements have scalariform thickenings, but this will be 
considered later. 

The mixed pith in this species does not play so conspicuous a part as in Ankyropteris 
corrugata and the other known Zygopterid stems. The radiating arms of parenchyma 
and tracheides so prominent in A. corrugata are certainly present in M. duplex, but 
they are not very well marked. In all cases they are closely connected with the 
emission of the petiole-trace, as, for example, the radial arm figured in PI. II. fig. 15, a, 
which is shown to be intimately related to one of the protoxylem groups of the trace 
of PI. I. fig. 7. Indeed, the insertion of the petiole-trace into the stem has probably 
been the cause of the stellate appearance of the latter in Etapteris di-upsilon (Zygo- 

* P. Bertrand, fitudes sur lafronde des ZygoptAidees, p. 206, Lille, 1909. 



184 



W. T. GORDON 



pteris G7'ayi) ; and a similar cause may be assigned to the apparent grouping of the 
xylem into certain areas in A. corrugata and A. Brongntarti. In other words, the 
departure of the petiole-trace is beginning to have a greater effect on the stem xylem 
in M. duplex than it had in Diplolabis, and the series from Diplolabis to Etapteris 
(Zygopteris Grayi), through Metaclepsydropsis duplex and Ankyropteris corrugata 
and Brongniarti, foreshadows a type of petiole-trace departure which will no longer be 
protostelic, but cause a gap in the outer xylem ring. In the Zygopteridese the departure 
seems to have been protostelic in all cases, but the Osmundacese show the change from 
the one type to the other. M, duplex is thus, as far as the axis is concerned, distinctly 
intermediate between the Diplolabis type and that of Ankyropteris corrugata ; indeed, 
it holds the same position among the Zygopteridese with four rows of primary pinnae, 
that Ankyropteris corrugata does among those with two orthostichies of such 
appendages. 

Dividing up the Zygopteridese according to this criterion, as Kidston and Gwynne- 
Vaughan have done, we establish the two stem series: (1) Diplolabis romeri — Meta- 
clepsydropsis duplex — Etapteris di-upsilon (Zygopteris Grayi), and (2) Ankyropteris 
corrugata — Ankyropteris scandens. To make the intermediate position of M. duplex 
in the first series quite clear, the following table has been inserted : — 





Diplolabis. 


Metaclepsydropsis. 


Etapteris. 


Shape of stern xylem in transverse 
section 


Circular, with inner 
and outer zones. 


Circular, with inner 
and outer zones. 


Pentagonal, with 
inner and outer 
zones. 


Type of stele .... 


Solid. 


With conjunctive 
parenchyma in 
central zone. 


With conjunctive 
parenchyma in 
central zone. 


Type of tracheide - 


' Outer zone . 
Inner ,, 


Long, pointed, reti- 
culate. 

Short, square-ended, 
reticulate. 


Long, pointed, re- 
ticulate. 

Long, pointed, reti- 
culate or scalari- 
form. 


Long, pointed, sca- 
lariform. 

Long, pointed, sca- 
lariform. 


Anns radiating from inner zone . 


Absent. 


Slightly developed. 


Strongly developed. 


Type of branching .... 


Equal dichotomy. 


Equal dichotomy. 


Unequal dichotomy, 
so-called "axil- 
lary" branch. 



It has been shown that the reticulate type of thickening on the walls of tracheides 
is more primitive than the scalariform type, so that as far as we can judge M. duplex 
occupies a position above Diplolabis in the Zygopterid series ; and the other criteria at 
our disposal all point to the same conclusion. 

Before passing to a discussion of the systematic position of the petiole, I wish to 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 185 

enumerate the reasons for believing that the stem was a rhizome : — (1) The xylem of the 
stem is small compared with that of the petiole ; (2) the internodes (if we may call the 
distance between two petioles by this name) are long ; (3) what evidence there is re- 
garding the distribution of adventitious roots seems to indicate that they are irregular 
in their occurrence ; (4) the petiole-trace is at first small and then grows larger, as 
though it were supported in the lower portion either by overlying vegetable matter or 
soil, and did not attain its maximum development until it got above the substratum ; 
(5) there is no sclerenchyma in the cortex of the stem or of the lowest part of the 
petiole ; (6) the pinnse are in four orthostichies on the petiole, and the latter must there- 
fore have been held erect. 

The study of the petiolar development has demonstrated certain points of interest. 
The term "arm" has sometimes been used to characterise the portion of the petiole- 
trace between the protoxylem groups and the " horizontal bar," and while such arms do 
exist in Diplolabis and most other Zygopterid petioles, they do not occur in Metaelepsy- 
dropsis duplex. The increasing size of the island of parenchyma at the emergence of 
a pinna-trace-bar from the petiole no doubt gives the petiole-trace at certain levels the 
appearance of having such arms (PI. II. fig. 18), but this stage is not constant. As has 
been already pointed out, the ends of the trace in M. duplex have become dilated 
instead of being produced into arms. This has permitted the insertion on the petiole 
of more robust pinnae. In Dineuron we find a petiole-trace quite similar to that of the 
early stages in M. duplex, but in no case have petioles of the former genus been dis- 
covered which had dilated ends. 

Etapteris, on the other hand, appears to be a case where the dilatations have increased 
to such an extent that they have become quite arm-like. These arm-like processes are 
quite different from the arms in Diplolabis ; they are mere swellings similar to the 
dilatations on the outer ends of the arms in Diplolabis and on the ends of the trace in 
M. duplex. Dr P. Bertrand has pointed this out in his memoir on the Zygopterid 
petiole-trace, and has also demonstrated the manner in which the pinna-traces depart from 
the petiole-trace. As Dr Bertrand has further pointed out, there is some slight trace 
of arms in Etapteris comparable with those in Diplolabis. This will be made clear by 
a glance at text-fig. 4. 

In order to decide whether arms are present on the petiole-trace or not, it is 
necessary to examine a section immediately above the level of a pinna-trace-bar 
departure. In the case of Diplolabis very distinct arms may be seen, with protoxylem 
groups at the ends, and the same applies to Etapteris ; but Metaelepsydropsis and 
Dineuron do not exhibit such arms, the ends of the trace have grooves in them, and in 
these grooves lie the protoxylem elements. The grooves of Metaelepsydropsis and 
Dineuron are therefore equivalent to the wide bay in Diplolabis, Zygopteris, and 
Etapteris. In all cases except Etapteris the development of new pinna-traces is 
essentially similar ; small tongues of xylem elements are developed round each proto- 
xylem group, and these grow towards one another until they meet and form a xylem 



186 W. T. GORDON 

bar across the end of the trace. The bay or the groove now becomes an island of 
parenchyma. 

In Diplolabis the xylem bar is well developed and the size of the island is constant, 
as is also the case in Zygopteris. In Etapteris the tongues of xylem do not meet, but 
break away and unite after becoming detached ; this is rather a specialised type of 
pinna-trace departure. Dineuron and Metaclepsydropsis, however, exhibit quite a 
different type. The small groove is bridged across much more quickly than in 
Diplolabis and Zygopteris, but subsequent growth causes the island to become larger 
and larger until it reaches a maximum at the level of the departure of the pinna-trace- 
bar. In Dineuron, it is true, the island does not reach the proportionate dimensions 
that it does in Metaclepsydropsis, but the increase is distinct, and I believe that these 
two genera must be grouped together. 

Such a grouping would necessitate the division into two groups of the first sub- 
division of the Zygopteridese with quadriseriate pinnae, as given by Kidston and Gwynne- 
Vaughan. In the first group would be included all forms with well-marked arms and 
the bay between them always constant in size : — 

Diplolabis rbmeri (Solms). 
Zygopteris primaria (Cotta). 

The second group would include forms where distinct arms are not developed, and 
where consequently the bay is reduced to a mere groove. The island of parenchyma 
formed by the bridging of this groove becomes gradually larger until a maximum is 
reached just before the departure of the pinna- trace-bar : — 

Dineuron ellipticum, Kidston, and D. pteroides, Renault. 
Metaclepsydropsis duplex (Williamson). 

This subdivision, however, is based entirely on the mature form of the petiole-trace. 
In certain members of each division [Diplolabis and M. duplex) it has been shown that 
in early stages of petiolar development the traces are distinctly similar in appearance 
to that of Dineuron. 

hi a recent paper on Diplolabis rbmeri* I drew attention to an hypothetical type 
of petiole-trace (Protoclepsydropsis) from which some other Zygopterid traces might be 
derived, and, in a tabular form, indicated what I believed to be the relationship of these 
Zygopterid petioles to one another. The table is inserted below (text-fig. 5), and in it 
.1/. duplex has been placed in close relation to Dineuron ; in fact, it has been considered 
one of the forms directly derived from a Dineuron ancestry. Zygopteris and Diplolabis 
have been grouped together as the second derived form, and Etapteris as the third. It 
is exceedingly interesting to find that the discovery of the stem and early stages of the 
petiole-trace of M. duplex has entirely confirmed the view set forth in that table. 

* Cordon, "On the Structure and Affinities of Diplolabis rovieri (Solms)," Trans. Boy. Soc. Edin., vol. xlvii. 
]it. iv. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 187 



Of course all this grouping has been based on the hypothesis that the forms of the 
trace at low levels in the petiole (i.e. at early stages of development) are comparable 
with ancestral forms. This is also the basis of the work of Kidston and Gwynne- 
Vaughan on the fossil Osmundaceae, and of Sinnott's studies on recent ferns. The 
hypothesis seems reasonable, and when applied to Diplolabis and Metaclepsydropsis 
(two forms closely allied on other grounds) the results lead to the same conclusion, 
and show that evolution was parallel in these two genera. 

As far as the relationships between the Osmundaceae and Zygopterideae are con- 
cerned, the stem of M. duplex occupies an important position. In the medullation 
of the Zygopterid stele the series taken was Diplolabis. Ankyropteris corrugata, 
Etapteris di-upsilon (Z. Grayi). Now this is open to criticism, for the first and last 
named have quadriseriate primary pinnae and petiole -traces with two planes of 



I 



1 



ProToclepsydropsis 



Clepsydropsis 



Dineuron 



MeTa clepsydropsis 




Efapferis 



Text-Fig. 5. — Table to show the relation of Mctaclepsydro2)sis to the other Zygopteridese which have quadriseriate pinnae. 

symmetry, while the central type has biseriate primary pinnae and one plane of 
symmetry in the petiole-trace. If we substitute M. duplex for Ankyropteris we 
strengthen the position from two points of view, for ( 1 ) we get a series of forms which are 
similar in all respects, and (2) we may construct a parallel series in the Zygopterideae with 
biseriate primary pinnae and one plane of symmetry, thus : — Ankyropteris corrugata — 
A. scandens. To complete the series in this second main division of the Zygopterideae 
we only want a form with a solid stem stele similar to Diplolabis. Clepsydropsis may 
supply this form, for the stems referred to that genus by Dr Paul Bertrand * are not 
above suspicion. Typical Clepsydropsis petioles have not been traced into these stems, 
and until this is done the Cladoxylons cannot be accepted as the stems of Clepsydropsis. 
Meanwhile the general trend of evolution as shown in the Zygopterideae is parallel 
to that demonstrated by Kidston and Gwynne-Vaughan in the Osmundaceae, namely, 
from a simple to a more complex type, i.e. both groups show an ascending series as 
we pass upwards in the geological succession. 



* Gomptes Rendus des Seances da I'Acade'mie des Sciences, Paris, 16th November 1!)08. 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 8). 



29 



188 W. T. GORDON 

In concluding this paper I desire to express my thanks to Dr Kidston, Dr Scott, 
and Miss Benson, with whom I have been in constant communication during my work 
on the genus described here. I have also benefited greatly by a study of Dr Paul 
Bertrand's memoir on the Zygopteridean frond, and, although I cannot agree with 
all his conclusions, this work has been of great service to me. 

I have also to express my thanks to the Executive Committee of the Carnegie 
Trust for defraying the expenses of illustrating this paper. 



EXPLANATION OF PLATES. 

{Photographs from untouched negatives.) 

Unless otherwise stated, the sections figured are in the author's collection. 

Plate I. 
Metaclepsydropsis duplex. 

Fig. 1. Transverse section of stem, x 1 = outer xylem ; x 2 = pith with tracheides; p = parenchyma. Slide 
1129. x 28. 

Fig. 2. Longitudinal section of stem, x 1 — outer xylem ; x 2 = inner xylem. Slide 1128. x 28. 

Fig. 3. Dichotomy of the stem, stage 1. a; 1 = outer xylem; x 2 — inner xylem; rt. tr = root trace. Slide 
1120. x 18. 

Fig. 4. Dichotomy of stem, stage 2. Lettering as in last figure. Slide 1121. x 18. 

Fig. 5. Dichotomy of stem, section above fig. 4. x 1 = outer xylem; .r 2 , x 2 = inner xylem. Slide 1122. 
x 18. 

Fig. 6. Dichotomy of stem just before the branches separate. Lettering as before. Slide 1123. x 18. 

Fig. 7. Insertion of petiole-trace in stem. Petiole-trace detached, prx 1 , prx 2 = protoxylem groups; 
x 1 and x 2 as before. Slide 1110. x 18. 

Fig. 8. Insertion of petiole-trace in stem. Petiole-trace attached, x 1 = outer zone of xylem; x 2 = inner 
zone ; prx 1 , prx 2 = protoxylem ; J = large xylem elements. Slide 1109. x 18. 

Fig. 9. Stage below that of fig. 8. prx 1 , prx 2 = protoxylem groups. Slide 1108. x 18. 

Fig. 10. Section further in than fig. 9. prx 1 , prx 2 = protoxylem groups as before. Slide 1107. x 18. 

Fig. 11. L'dow fig. 10. a 1 = radial arm from inner xylem zone; prx 1 , prx? = protoxylem groups. Slide 
1106. x 18. 

Fig. 12. Stage preceding that of fig. 11. Lettering as before. Slide 1105. x 18. 

Plate II. 
Metaclepsydropsis duplex. 

Fig. 13. Insertion of petiole-trace in stem. Stage below PI. I. fig. 12. prx 1 , prx 2 = protoxylem groups ; 
a 1 =arm of parenchyma and tracheides radiating from inner xylem zone. Slide 1104. x 18. 

Fig. 14. Section following that of fig. 13. prx 1 = first protoxylem group, a 1 = sinus which is really 
the beginning of another radial arm similar to that of fig. 16. Slide 1103. x 18. 

Fig. 15. Section below that shown in fig. 14. prx} = first protoxylem group of petiole-trace; a = radial 
arm of parenchyma and tracheides. Slide 1102. x 18. 

Fig. 16. Section below that shown in fig. 15. prx 1 = protoxylem group"; a = arm of tracheides and 
parenchyma radiating from inner zone of stem axis. Slide 1101. x 18. 

Fig. 17. Transverse section of mature petiole-trace showing hour-glass shape. w = waist; 6 = pinna-trace- 
bar; ,'/ = groove; prx = protoxylem groups ; end = endodermis. Slide 1146. x 17. 



ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 189 

Fig. 18. Transverse section of mature petiole-trace showing a large island (is) of parenchyma at one end 
and an open groove at the other. w = waist; b = pinna-trace-bar ; g = groove; prx = protoxylem group; 
is = island of parenchyma; c = shallow depression — the last vestige of the groove. Slide 1155. x 17. 

Fig. 19. Section similar to fig. 18. Pinna-trace-bar just detached from petiole-trace. Waist not so well 
marked as in last two figures. Lettering as before. Slide 1151. x 17. 

Fig. 20. Example of a petiole-trace with waist still less marked than in fig. 19. At the top an early 
stage in the development of the pinna- trace-bar is shown. Lettering as before. Slide 1167 x 17. 

Fig. 21. Transverse section of petiole-trace in which the waist is scarcely developed at all. A pinna- 
trace-bar is shown attached to the petiole-trace at the top of the figure. Pra J \, prx b , prx\, prx° 2 = proto- 
xylem groups on petiole-trace; prx v prx. 2 , prx 3 , prx± = protoxylem groups on pinna-trace-bar; is = island of 
parenchyma; b = pinna-trace-bar. Slide 1171. x 17. 

Fig. 22. Section of a petiole-trace from which all indication of a waist has vanished. Lettering as in 
other figures. Slide 1152. x 17. 

Fig. 23. In the petiole-trace of this figure a further reduction to an elliptical trace is clearly seen. The 
pinna- trace-bar (b) is a much reduced one, and the pinna-traces do not pass beyond the cortex of the 
petiole, i.e. they are reduced. Slide 1264. x 165. 

Fig. 24. Transverse section of petiole-trace showing distinct elliptical form. Pinna-traces in this case 
also are reduced. The outer cortex (o.c) has no sclerotic layer. Slide 1194. x 18. 

Fig. 25. Rather oblique transverse section of a reduced petiole-trace. Pinna-trace-bar (b) does not 
divide into two pinna-traces. Slide 1189. x 15. 

Fig. 26. Another oblique transverse section of petiole-trace. Pinna-trace-bar shows a double curve, but 
it does not divide into two to supply two pinnae. Slide 1204. x 15. 



Plate III. 

Metaclepsydropsis < lup/e.r. 

Fig. 27. Transverse section of petiole-trace. Fig. 26 shows this same trace at a higher level. Slide 
1199. xl5. 

Fig. 28. Transverse section of petiole-trace similar to that of fig. 27. Slide 1184. x 15. 

Fig. 29. Transverse section of another closely similar petiole-trace, prx = protoxylem groups ; g = groove. 
Slide 1189. x 17. 

Fig. 30. Transverse section of petiole-trace similar to that of rig. 29. It is inserted to show resemblance 
at this level of the traces of Metaclepsydropsis and Dineuron. Slide 1180. x 16"5. 

Fig. 31. Transverse section of petiole with trace slightly flattened. The outer cortex contains very 
little sclerenchyma, if any. Slide 1252. x 16 - 5. 

Fig. 32. Furthest-out section of a series showing the connection between petiole-trace and stem xylem. 
Slide 1242. x 15'5. 

Fig. 33. Section below that shown in fig. 30. The interval between is about ]- inch, prx, prx = proto- 
xylem groups ; is = island of parenchyma ; st = xylem of stem. Slide 1240. x 15 - 5. 

Fig. 34. Section some distance below that of fig. 31. Petiole-trace crushed against stem. Two sections 
below this the two unite, but are very much crushed at that level. Slide 1233. x 155. 

Fig. 35. Transverse section of petiole to show the arrangement of the tissues. The sclerotic outer zone 
of the cortex is very distinct. The petiole-trace is shown more highly magnified in PI. II. fig. 21. pet. tr = 
petiole-trace; is 1 , is 2 — islands of parenchyma; sc.o.c = sclerotic outer cortex; o.c = parenchymatous outer 
cortex. Slide 1171. x 8-3. 

Fig. 36. Transverse section of petiole-trace at a high level in the petiole. Slide 1140. x 17. 

Fig. 37. Transverse section of petiole-trace and combined pinna- and aphlebia-traces. is = island of 
parenchyma; a 1 , a 1 = combined pinna- and aphlebia-traces ; prx = protoxylem groups. Slide 1248. x 19. 

Fig. 38. Transverse section above that of fig. 37. The combined traces show signs of a division into two ; 
the larger (upper) trace is the pinna-trace; pin. tr = pinna-trace ; aph. tr = aphlebia-trace ; £>?•# = proto- 
xylems. Slide 1244. x 19. 



190 ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 

Plate IV. 
Metaclepsydropsis duplex. 

Fig. 39. Transverse section of petiole-trace above level of PI. III. fig. 38. <j = groove; pin. tr = pinna- 
trace ; aph. tr = aphlebia-trace. Slide 1243. x 19. 

Fig. 40. Transverse section petiole to show the various traces departing from it. pet. tr = petiole- 
trace ; pin. tr = pinna-traces ; aph. tr = aphlebia-traces. Kidston Collection, 1314. x 3"2. 

Fig. 41. Transverse section of normal primary pinna-trace. The shape is like a horse-shoe with ends 
incurved. Slide 1261. x 40. 

Fig. 42. Transverse section of an abnormal secondary pinna-trace with tertiary pinna-trace departing 
from it. Both traces are closed rings. Slide 1258. x 40. 

Fig. 43. Transverse section of abnormal primary pinna-trace. The trace is a closed ring of xylem. 
Slide 1168. x40. 

Fig. 44. Transverse section of root-trace showing that it is diarch. pra = proto xylem group. Slide 
1101. x 18. 

Fig. 45. Transverse section of stem to show inner tracheides and parenchyma, x 1 = outer xylem; 
a; 2 = inner zone ; p = conjunctive parenchyma. Slide 1109. x 36. 

Fig. 46. Transverse section of branch resembling Botrychioxylon. In places the xylem shows a perfect 
radial arrangement. Slide 1233. x 18. 

Fig. 47. Transverse section of stem with petiole-trace attached. The stem has a central zone almost 
entirely composed of parenchyma ; only one tracheide can be seen. The petiole-trace is much distorted. 
p = parenchyma in the stem ; t = tracheide. Slide 1207. x 15. 



BIBLIOGRAPHY. 

(1) Arbbr, E. A. N. "On the Past History of Ferns," Ann. Bot, vol. xx, July 1906. 

(2) Bertrand, P., Etudes sur lafronde des Zygopteridees, Lille, 1909. 

(3) Bower, F. O., On the Origin of a Land Flora, London, 1908. 

(4) Corda, Beitrcige zur Flora der Vorwelt, 1845. 

(5) Cotta, B., Die Dendrolithen, 1832. 

(6) Gordon, W. T., "On the Structure and Affinities of Diplolabis romeri (Solms)," Trans. Roy. Soc. Edin., 

vol. xlvii., pt. iv., 1911. 

(7) Kidstox, R., and Gwynne-Vaughan, D. T., "On the Fossil OsmundaceaV' Parts I.-IV., Trans. Roy- 

Soc. Edin., vol. xlv., pt. iii., 1907; vol. xlvi., pt. ii., 1908; vol. xlvi., pt. iii., 
1909; vol. xlvii., pt. iii., 1910. 
,, ,, " On the Origin of the Adaxially Curved Leaf-trace in the Filicales," Proc. Roy. Soc. 

Edin., vol. xxviii., pt. iv., 1908. 

(8) Renault, B, "Etude sur quelques vegetaux silicifies d'Autun," Ann. des Sciences naturelles, ii. 

serie, Botanique, xii., 1869. 
,, ,, "Bassin houiller et permien d'Autun et d'Epinac," Flore Fossile : OUes minerau.r de 

la France, 1896. 

(9) Richteu u. Unger, "Beitrag zur Paheontologie des Thuringer Waldes," Denkschr. d. K. A'. 

Alcademie zu Wien, Math.-Naturw. CI., Band xi., 1856. 

(10) Scott, D. H., Studies in Fossil Botany, 2nd edition, vol. i., 1908. 

(11) Seward, A. C, Fossil Botany, vol. ii., Cambridge, 1910. 

(12) Sinnott, E. W., "The Evolution of the Filicinean Leaf-trace," Ann, Bot., vol. xxv., 1911. 

(13) Solmh-Laubach, H. Graf zu, " Ueber d. in d. Kalks. d. Kulm v. Glatzisch-Falkenberg in S. erhalt 

Structurb. Pflanzenreste," Botan. Zeitung, vol. 1., 1892. 

(14) Stenzel, G., "Die Oattung Tubicaulis Cotta," Mitth. aus dem Kgl. min. geol. Museum in Dresden, 

Heft viii., 1889. 

(15) Tanslby, A. G., "Lectures on the Evolution of the Filicinean Vascular System," New Phytologist, 1907. 

(16) Williamson, W. C, "On tin' Organisation of the Fossil Plants of the Coal Measures," Phil. Trans. 

Roy. Soc, vol. clxiv., 1871. 



Trans. Roy. Soc. Edin r Vol. XLVIII. 

W. T. Gordon : On the Structure and Affinities of Metaclepsydropsis duplex. — Plate I 





prx' 




prx 1 



Gordon. Photomic. 



M'Karlane & Erskine. Liti) , Edin 



Trans. Roy. Soc. Edin r y l XLVIII. 

W. T. Gordon : On the Structure and Affinities of Metaclepsydropsis duplex.— Plate II, 



a.' 'prx 




r<lon. Photomif. 



M t'arlanc k Eiskinc, Lit.h.. Edin 



Trans. Eoy. Soc. Edin r Vol. XLV1II. 

W. T. Gordon : On the Structure and Affinities of Metaclepsydropsis duplex. — Plate III. 







prx 



V T. Gordon. Photomic. 



M'Farlane & Erskirje, Lith., Edin. 



Trans. Roy. Soc. Edin r Vol. XLVIII. 

W. T. Gordon : On the Structure and Affinities of Metaclepsydropsis duplex. — Plate IV 




T. Gorlon. Photouiic. 



M Farlane & Erekine. F/Uli Edir 



( 191 ) 



IX. — Scottish National Antarctic Expedition: Observations on the Anatomy of 
the Weddell Seal (Leptonychotes Weddelli). By David Hepburn, M.D., CM., 
Professor of Anatomy, University College, Cardiff (University of Wales). 
Part II * 

(MS. received December 4, 1911. Read January 8, 1912. Issued separately January 19, 1912.) 

Genito-urinary Organs. 

In my former contribution I gave a general summary of the animal under considera- 
tion, and discussed in detail the peritoneal arrangements of its abdominal cavity and 
the naked-eye anatomy of its alimentary organs. In the present paper I shall give an 
account of the genito-urinary system. 

The kidneys were situated on each side of the dorsal mesial mesentery. Each was 
covered on its ventral aspect by the peritoneum forming the dorsal wall of the greater 
peritoneal sac. The right kidney was quite free from contact with the liver and the 
duodenum, while the left kidney was equally free from contact with the spleen. Both 
kidneys were therefore situated well back towards the pelvic end of the abdominal 
cavity. Each kidney measured 5 inches in the longitudinal diameter and 2 inches 
in the transverse diameter. The hinder or caudal end of each reached a point two 
inches from the pelvic inlet, which, as formerly described, was narrow and well defined 
by the course of the hypogastric (umbilical) arteries. 

The surface of the kidney indicated lobulation, but the lobules were not separated from 
each other. The hilum was placed ventro-mesially, and at its point of emergence from the 
surface of the kidney the ureter was nearer to the caudal than to the cephalic end of the 
organ. On opening up the hilum, the ureter was seen to result from the union of two main 
tributaries, each of which, in its turn, was formed by the junction of several smaller root- 
lets, which corresponded more or less closely in number to the number of the kidney 
lobules. There was no distinct pelvis to the ureter, which was gradually formed by the 
junction of smaller ducts in the manner indicated. Nevertheless, the widest point of 
the ureter was found at the junction of its two main tributaries. The ureter and its 
chief tributaries lay on the ventral (anterior) aspect of the renal vessels, and not on their 
dorsal (posterior) aspect, as is the case in man. 

The size of the ureter suggested a vessel about half the diameter of an average 
human radial artery. The ureter followed a course along the dorsal wall of the abdomen 
towards the pelvic inlet ; and half an inch beyond the termination of the abdominal 
aorta, or, in other words, at the point where the common iliac artery divided into its 
external and internal branches, the ureter crossed to the mesial side of the internal iliac 
and hypogastric arteries, and continued its course along the margin of the pelvic inlet. 
In this position the ureter and the hypogastric artery were both in their turn crossed 

* Part I. was published in the Trans. Roy. Soc. Edin., vol. xlvii., pt. i. (No. 3), 1909. 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 9). 30 



192 PROF. DAVID HEPBURN 

by the vas deferens, which now assumed the mesial position to both of the others. Up to 
this point the ureter had not entered the pelvic cavity, and about three-quarters of an 
inch onwards, that is, in the direction of the tail, the ureter, still lying close to the pelvic 
brim, entered the lateral aspect of the urinary bladder, travelling between the folds of 
a lateral vesical mesentery or peritoneal ligament. Thus, as a consequence of the great 
obliquity of the pelvic inlet, the ureter was able to reach the bladder by skirting the 
pelvic brim, and at no point did it require to enter or sink into the interior of the pelvis. 

The urinary bladder was placed mesially, and was attached to the ventral wall of 
the abdomen by a ventral mesial mesentery composed of peritoneum, which, as formerly 
described, closely invested the bladder except on its pubic aspect. The apex of the 
bladder extended to the umbilicus, where it still presented an open lumen. There was 
no obliterated part or urachus, and throughout its entire length it presented a uniform 
calibre, suggestive of an empty portion of small intestine. Developmentally, it may be 
said to represent an enlarged and patent allantois ; but as this animal was only two days 
old at the time of its death, probably a sufficient period had not elapsed for the closure 
of the umbilical end of the organ. 

The hypogastric arteries were carried along each lateral aspect of the bladder, 
suspended in peritoneal folds half an inch in width, so that these arteries were not in 
contact with the wall of the bladder until they reached a point between 2 and 3 inches 
from the umbilicus, where the peritoneal folds disappeared, and the arteries closed in 
upon the sides of the bladder. 

The length of the bladder from the umbilicus to the prostate gland was 10 inches. 
The prostate gland lay close to and on the abdominal side of the symphysis pubis. 

The interior of the bladder was lined by a mucous membrane, presenting numerous 
rugosities, which to a large extent lay parallel to each other, and in the longitudinal 
axis of the bladder. Towards the outlet the mucous membrane became comparatively 
flat and smooth. 

The orifices of the ureters were longitudinal oblique narrow slits 2 mm. in length and 
5 mm. apart.' The lateral margins of each of these openings were continued towards 
the outlet as slight ridges for a distance of 10 mm. These ridges met in the mesial, 
plane, thus forming a mesial longitudinal ridge or uvula vesicae. The actual trigonum 
vesicae was therefore a triangular area 5 mm. wide at its base and 10 mm. long on 
each side. 

The uvula vesicae was continued into the urethra, and became continuous with the 
crista urethrae, which attained its greatest prominence 20 mm. from the apex of the 
trigonum vesicae. The sinus pocularis was represented by a very small mesial aperture 
opening on the distal side of the summit of the crista urethrae. 

The prostate gland did not attract attention, and at first sight one would have 
doubted its presence. Certainly in cutting into the urethra from its pubic aspect no 
variation in consistence was detected. Still, there was a definite thickening of the pubic 
wall of the urethra corresponding to the general position of the urethral crest. On the 






ON THE ANATOMY OF THE WEDDELL SEAL. 193 

other hand, on the rectal or pelvic aspect of the urethra, and in relation to the urethral 
crest, there was a mesial longitudinal thickening of firm consistence, from 5-6 mm. in 
length. Into the hinder end of this denser part the vasa deferentia entered. The 
pelvic portion of the urethra was therefore not surrounded by visible prostatic tissue at 
its vesical end, and the prostatic tissue was not prolonged in relation to the urethra as 
far as the sub-pubic pelvic wall, because, whereas the prostate was only from 5-6 mm. 
in length, the pelvic urethra measured from 35-40 mm. long. No doubt the extreme 
youth of the animal accounts for the primitive condition of the prostate, but it is 
interesting to note that the part readily recognisable is the mesial longitudinal lobe. 
A portion of the prostatic part of the urethra, along with the surrounding tissue, was 
prepared for microscopic examination. Definite glandular prostatic tissue was revealed 
in relation to the pubic and lateral aspects of the urethra. On the rectal aspect of the 
urethra dense fibrous tissue was displayed. The two vasa deferentia were visible, each 
quite distinct from the other, so that their close proximity and apparent fusion previous 
to their entering the prostate on its rectal aspect was not a real fusion. The urethral 
crest presented the section faces of the bifurcated end of the uterus masculinus 
(Mullerian ducts). 

Each testis was lodged in its own peritoneal pouch, which was situated to the outer 
side of the pubic body, in the depression between the pubis and the head of the tibia. 
These pouches were completely separated from each other by the keel-like projection 
of the pubic symphysis, and thus they did not form any object comparable to a 
scrotum. The testes had descended through the abdominal wall on the ventral aspect 
of Poupart's ligament, i.e. through the inguinal canal, and not through the crural canal, 
notwithstanding the novel position occupied by the testis in relation to the limb as a 
whole. The tunica vaginalis testis was in open communication with the sac or cavity 
of the abdominal peritoneum, and thus the whole condition might fairly be said to 
resemble two imperfectly descended testes, although in this case there was no scrotum 
into which they could have descended, nor was it possible for them to descend any 
farther. Each testis was considerably flattened, and measured 25 mm. long by 14 mm. 
wide. 

No hydatids of Morgagni were visible. The epididymis presented a globus major 
and a globus minor. It did not lie close to the testis, but was supported by a mesentery, 
which at its deepest measured 10 mm. The vas deferens was similarly supported, and 
therefore it presented itself clear from the epididymis at the distal end of the testis, 
instead of lying close to it as far as the proximal end of the testis, as in man. 

The vas entered the inguinal canal in the usual way and crossed the iliac fossa, 
running superficial to the external iliac vessels and the hypogastric artery. Thereafter 
it hooked round the hypogastric artery, and, passing to its mesial side, it proceeded 
backwards, i.e. tailwards, towards the base of the urinary bladder, taking its place to the 
mesial side of the ureter on its course. As the vas approached the proximal or pelvic 
end of the prostate gland it came into such close contact with its fellow of the opposite 



194 PROF. DAVID HEPBURN ON THE ANATOMY OF THE WE UDELL SEAL. 

side that their adjacent walls became firmly blended together. This produced the 
appearance of an enlargement common to both of them, but there was no dilatation or 
ampulla on each one. There was no trace of seminal vesicles. 

The penis was constructed on familiar lines. A strong, flexible cylindrical structure 
was present in the body of the penis extending 7 mm. from the base of the glans 
penis backwards. A portion of this structure was removed for microscopical examin- 
ation. In transverse section it presented a circular outline, and was equally associated 
with the two corpora cavernosa penis. Its resistance to the knife suggested young 
bony tissue, and accordingly it was decalcified. Afterwards sections were cut out of 
paraffin, mounted, and stained in hsematoxylin and eosin. Under the microscope it 
presented the distinctive characters of cancellated bone, being more spongy towards the 
centre of the section and denser towards the surface, where it was closely enveloped in 
a fibro- vascular sheet of membrane, comparable to periosteum. Numerous bone-cells 
were embedded in the developing processes of bone. No trace of hyaline cartilage could 
be detected. No doubt this short cylindrical piece of young bone is comparable to the 
much larger os penis of the walrus, as well as to the furrow-shaped and partly bilateral 
os joenis of the fox and the dog. The bulb on each corpus cavernosum penis was situated 
in relation to the crus penis, and not on the penile portion of the organ. From the region 
between the bulb of the corpus spongiosum penis and the rectum, i.e. corresponding to 
the central point of the perineum, there were two parallel bands of tissue running forwards 
towards the distal end of the body of the penis. These were similar to muscular bands 
which I have elsewhere* described in connection with the penis of the porpoise. 
Probably these act as retractors of the penis. As in the case of the porjDoise, a 
microscopic examination of sections cut longitudinally and stained after Van Giesen's 
method revealed unstriped muscular fibres, with fibrous tissue bundles. Since there 
is no scrotum in the porpoise, whose testes are situated intra-abdominal, and since in the 
seal under consideration each testis occupied a recess placed under the integumentary 
layers in relation to the inner side of the head of the tibia, it seems not unfair to 
consider these non-striped muscular bands as being homologous to the tunica dartos 
layer of an ordinary scrotum, more especially as the muscular fibres of the tunica 
dartos are of the unstriped or involuntary variety. 

* " The Anatomy of the Genito-urinary Apparatus of the adult male Porpoise," Hepburn and Waterston, Trans. 
Royal Physical Society, Edinburgh, 1902. 



( 195 ) 



X. — The Influence of the Ratio of Width to Thickness upon the Apparent 
Strength and Ductility of Flat Test-bars of Mild Steel. By W. Gordon, B.Sc, 
A.M.I.Mech.E., Lecturer in Mechanical Engineering in Leith Technical College, 
and G. H. Gulliver, B.Sc, A.M.I.Mech.E., Lecturer in Engineering in the University 
of Edinburgh. 

(MS. received October 24, 1911. Read November 20, 1911. Issued separately February 28, 1912.) 

1. Introduction. 

In a large class of engineering structures it is essential that the materials employed 
should be both strong and ductile, so that not only shall the structure be able to resist 
heavy loads, but that if by any chance it is overloaded it shall not collapse suddenly. 
In order to ascertain whether a metal is suitable for a particular structure, its strength 
and ductility are determined experimentally. The test most commonly in use con- 
sists in applying a gradually increasing pull to a bar of the metal until fracture takes 
place. The maximum load supported per unit of the original cross-sectional area of the 
bar is called the tensile strength or tenacity of the metal, and the elongation of an 
initial measured length, expressed as a proportion of that length, is called the extension, 
and is used as an index of the ductility of the metal. 

Experience shows that both strength and ductility, as measured in the tensile test, 
depend not only upon the properties of the metal, but also upon the shape of the test- 
bar. If there are abrupt variations in the cross-sectional dimensions, the apparent 
strength of the metal is greater, and the apparent ductility is less, than when these 
dimensions are constant or vary gradually. Again, the apparent ductility diminishes 
as the original length of the test-bar is made greater, on account of the well-known 
phenomenon of constriction ; the bar suffers a considerable reduction of section in the 
region where fracture eventually takes place, and the extension in this part of the 
bar is correspondingly greater than elsewhere. It follows that the extensions of two 
bars of the same cross-sectional dimensions are not comparable unless the datum- 
lengths of the two bars are the same. 

In measuring the extension of bars, similar in form but differing in dimensions, 
comparable results are obtained only by observing Barba's principle of similitude, 
which may be stated thus: "Similar bodies of the same material remain similar 
when distorted by similar systems of applied loads." In other words, if the linear 
cross-sectional dimensions of one test-bar are double those of another, the datum- 
length of the former must be twice that of the latter. In practice, it is a great 
convenience to use test-bars of constant gauge-length, and this necessitates that the 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 10). 31 



196 



MR W. GORDON AND MR G. H. GULLIVER ON 



cross-sectional dimensions of all test-bars shall be the same also. This rule is observed 
generally in important work, but there are different standard gauge-lengths to suit 
different classes of material. The most common lengths in use in this country are 
8 inches and 2 inches (1), corresponding closely with the Continental lengths of 200 
millimetres and 50 millimetres. 

There is one remaining difficulty, namely, that of securing comparable measure- 
ments with bars of different sectional form. The importance of this point is such that 
the proportions of the test-bar may make all the difference between an apparently 
satisfactory and an apparently unsatisfactory material. The great majority of bars 
submitted to the tensile test are either of circular or of rectangular section, and the 
difficulty consists in comparing among themselves the values obtained from rectangular 
bars having varying widths and thicknesses, and in turn comparing these with the 
values obtained from round bars. The object of the experiments to be described was 
to determine the variation of apparent strength and ductility caused by variation in 
the width of soft steel test-bars of constant thickness. 



2. Earlier Investigations. 

The earliest experiments which bear upon the point under discussion appear to be 
those of Barba (2). This investigator cut from a steel plate, 10 mm. (0"39 inch) 
thick, a number of test-bars with widths of 10 to 80 mm., and obtained the results 
set forth in Table I. 

TABLE I. 
Tensile Tests of Soft Steel Babs 10 mm. Thick (Barba). 







Yield- 


point. 


Tenacity. 


i 
Extension per cent, on 


Width. 


Ratio, 
width 


























thickness ' 


Kg. per sq. 
mm. 


Tons per sq. 
inch. 


Kg. per sq. 
mm. 


Tons per sq. 
inch. 


50 mm. 


100 mm. 


10 


1 


24-8 


15-8 


38-4 


24-4 


37-6 


310 


20 


2 


24-6 


156 


401 


25-5 


45-0 


34-0 


30 


3 


25-4 


1GT 


39-4 


25-0 


48-0 


35-0 


40 


4 


25-0 


15-9 


39-8 


253 


52-0 


37-0 


50 


5 


24-6 


15-6 


38-1 


24-2 


560 


390 


60 


6 


24-9 


15-8 


377 


239 


61-0 


408 


70 


7 


24-8 


15-8 


37-8 


24-0 


57-0 


38-5 


80 


8 


235 


149 


38-4 


24-4 


52-0 


345 



The yield-point and the tenacity of the metal remain sensibly constant, but the 
extension, measured on two different gauge-lengths of 50 and 100 mm. respectively, 
shows considerable variation. In fig. 1 the extension is plotted against the ratio 
width/thickness, and both curves exhibit a well-marked maximum when the width is 
about six times the thickness. 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



197 



Experiments by investigators subsequent to Barba have thrown some doubt upon 
the existence of such a maximum. Appleby (3) found that the extension increased 
continuously up to the ratio width/thickness = 8, the highest ratio employed in his 
experiments, but he used a hard grade of Bessemer steel. The results obtained by 



70 



60 



50 



£40 



z 
o 

CO 

5 30 

t- 

X 



20 



10- 




4 
WIDTH 



THICKNESS 



Fig. 1. — Variation of extension of flat test-pieces with change in the ratio of width to thickness ; measurements obtained 
from bars of soft steel having widths from 10 to 80 mm., a constant thickness of 10 mm., and gauge-lengths of 50 and 
100 mm. (Barba). 

Unwin (4) with several different grades of mild steel and a constant gauge-length of 
8 inches, show frequently a maximum value of the extension, but the ratio of width 
to thickness at which this maximum occurs is variable and often ill-defined. More- 
over, in most of the series of test-bars the thickness was varied. This method has the 
disadvantage, from the present point of view, that it is difficult with a variable 
thickness of plate to maintain a sufficiently uniform quality of material. 

Consideration of the principle of similitude suggests that the extensions of bars of 
dissimilar section would be more nearly constant if, instead of a fixed gauge-length, 



198 MR W. GORDON AND MR G. H. GULLIVER ON 

one proportional to the square root of the cross-sectional area were employed. By 
means of tests of round, square, flat, angle, and channel bars, Barba has shown that 
this expectation is nearly fulfilled (2). The longer standard round test-bar in use on the 
Continent is 20 mm. diameter and 200 mm. long, the length being thus H'3-s/area; 
bars of other shapes and sizes are therefore given a gauge-length of 11'3^/area when 
comparable measurements are required. The longer British standard round test-bar 
has a length eight times its diameter, equivalent to 9 ^/area ; the shorter round 
standard has a length of 4^/area (1), but for bars of rectangular section the length is 
fixed at 8 inches. 

So long as the section is of compact form — that is, circular, square, or rectangular 
with a width not greater than about three thicknesses — the extension of any bar 
conforms closely with a simple linear equation clue to Unwin (4) : — 

e per cent. = \Oofa + b ■ ^ A \ , 

in which a and b are constants, A is the cross-sectional area, and L is the datum- 
length. In the case of rectangular bars having a breadth considerably exceeding the 
thickness, this simple equation is not fulfilled. 



3. Experimental Details. 

The primary object of the present experiments is to investigate the variation in 
the extension of flat bars of mild steel of constant thickness as the width is gradually 
increased. The advantage of varying the width instead of the thickness of the test- 
bars is twofold : — 

(1) The work of preparing the bars is less. 

(2) The material is of more uniform quality. 

The material employed was soft boiler plate, \ inch thick. No chemical analysis 
was made, but a section of the metal, magnified 100 diameters, is given in fig. 2. 
This shows the usual features of soft steel — the ground mass of ferrite or a-iron, the 
dark areas of pearlite or Fe-Fe 3 C eutectoid, and a few small slag inclusions of elongated 
form, — and it indicates the presence of 0'12 to 0"15 per cent, of carbon. 

The plate was cut, in the direction of rolling, into strips, approximately ^, 1, 1}, 
2, 1\, 3, 3^, and 4 inches wide, and these were machined carefully on the edges, and 
lightly ground on the wide faces to remove the mill scale. There were three bars of 
each width, with the exception of those 3} 2 inches wide, of which there were only two. 
Each bar had a total length of 18 inches, of which 3 inches at each end was held 
within the grips of the testing machine. A centre ]ine was scribed on both wide 
faces of each bar ; that on one face was divided carefully into inch lengths, and on the 
other a length equal to 1 1 - 3 times the square root of the area of the bar was set out 
and subdivided into a number of equal parts. On both faces the divisions were 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



199 



continued over the middle 12 inches of the length, in order that the extension might 
be computed in a satisfactory manner. The width and thickness of each bar were 
determined by a micrometer to "001 inch ; the maximum variation in dimensions did 




Fig. 2. — Longitudinal section of soft steel plate containing - 12 to 0*15 per cent, of 
carbon ; etched with picric acid and sodium picratc ; magnified 100 diameters. 

not exceed 0*5 per cent. In Table II. are given, averaged for each set of bars, the 
cross-sectional dimensions and area, the square root of the area, and the ratio of 
width to thickness. 

TABLE II. 

Dimensions of Test-bars. 



Nominal Width. 


Actual Width. 


Thickness. 


Area. 




Width 


^/Area. 


Inches. 


Inches. 


Inch. 


Sq. inches. 


Thickness 


i 


0-450 


0-255 


0-115 


0-339 


1-76 


1 


0-967 


0-254 


0-245 


0-495 


3-81 


1* 


1-443 


0-256 


0-370 


0-608 


5-64 


2 


1-947 


0-257 


0-501 


0-708 


7-57 


2 A 

2 


2-478 


0-259 


0-642 


0-801 


9-57 


3 


2-897 


0-260 


0-752 


0-867 


11-14 


H 


3-451 


0-259 


0-894 


0-945 


13-32 


4 


3-955 


0-257 


1-016 


1-008 


15-39 



The bars from ^ inch to 3 inches wide were tested in the Riehle (gear-driven) 
machine, and those 3^- and 4 inches wide in the Buckton (hydraulic) machine of the 
Engineering Department of the University of Edinburgh. When setting the bars in 
the testing machine, the centre line of one face was plumbed so that the load should be 
as nearly axial as possible. The wedge-grips of the Riehle machine, in which most of 



200 



MR W. CORDON AND MR G. H. GULLIVER ON 



the bars were tested, are made with slightly convex faces to assist in this respect. 
The rate of straining was kept constant throughout the tests at about 0i2 inch per 
minute, that is, I per cent, per minute on the 12-inch length between the grips. 

The loads noted during the test were the yield load, at which permanent stretch 
becomes well marked, the maximum load carried, and the load sustained by the bar at 
the moment before rupture ; the last is difficult to determine with great accuracy. 
After fracture the various dimensions of each bar were re-measured for comparison with 
the original dimensions. From these measurements the variations in mechanical 
properties, corresponding with change in the ratio of original width/original thickness 
of the bars, have been deduced. The more important quantities dealt with are : — 

(1) The yield point, the tenacity, and the mean breaking stress. 

(2) The extension on various constant gauge-lengths, on gauge-lengths proportional 
to the ratio width/thickness, and on gauge-lengths proportional to the square root of 
the cross-sectional area. 

(3) The reduction of area at fracture. 

4. Yield Point and Tenacity. 
The yield point is the stress at which there is marked evidence of permanent 
distortion of the metal ; it is found by dividing the yield load by the original area of 
cross-section of the bar. From the figures eriven in Table III. it is evident that the 
yield point is not affected sensibly by a change in the ratio width/thickness. The 
average yield point is at 18*14 tons per square inch. 

TABLE III. 

Variation of Yield Point, Tenacity, and Mean Breaking Stress, with the Ratio 

Width/Thickness. 



Nominal Width. 
Inches. 


Width 
Thickness 


Yield Point. 
Tons per sq. inch. 


Tenacity. 
Tons per sq. inch. 


Mean Breaking 

Stress. 

Tons per sq. inch. 


h 


1-76 


18-17 


25-26 


54-29 


1 


3-81 


18-28 


25-32 


. 50-64 


H 


5-64 


18-29 


25-45 


48-77 


2 


7-57 


18-10 


25-45 


46-00 


H 


9-57 


IS -08 


25-47 


4987 


3 


11-14 


18-00 


25-40 


47-91 


H 


13-32 


18-05 


25-66 


50-03 


4 


15-39 


18-16 


25-87 


50-75 



The tenacity is the maximum load carried by the bar divided by the original cross- 
sectional area. This quantity also remains almost constant throughout the series of 
bars, though there is a slight tendency to rise as the width is increased ; the figures 
are given in Table III. The average tenacity is 25 '49 tons per square inch. 

The tenacity, as defined above, though a useful quantity to the engineer, furnishes 
no information as to the maximum stress sustained by the metal. A near approxima- 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



201 



tion to the actual maximum stress is obtained by dividing the actual load supported at 
the instant before fracture by the minimum cross-sectional area of the broken bar ; it 
may be called the mean breaking stress. The actual load carried just before rupture 
is difficult to determine with accuracy on account of the localisation of the distortion 
and the relative rapidity with which the metal extends in this region, but much care 
was taken to secure correct results. In order to determine the minimum cross-sectional 
area of the broken bar the average thickness of the metal was obtained by a number of 
measurements at equidistant points along the fracture, and this was multiplied by the 
minimum width of the bar ; the difficulties of measurement do not admit of extreme 
accuracy here. The values of the mean breaking stress are given in Table III., and are 



bU 




















z 




















o- 
to 




















5 50 
















o 























Q. 








o 








z 
o 








o 












40 

in 
to 

UJ 

a. 

t- 

3D 







































8 10 

WIDTH 

THICKNESS 



12 



14 



16 



18 



Fig. 3. — Variation of mean breaking stress of flat test- pieces with change in the ratio of width to thickness ; measurements 
obtained from bars of soft steel having widths from | inch to 4 inches, and a constant thickness of J inch. 

plotted in fig. 3 ; the amount of variation is not great, hardly more than may be due to 
uncertainties of measurement. 

5. Extension. 

The measurements of extension were taken chiefly from that side of each bar which 
was divided originally into inch lengths, and were always made symmetrically on each 
side of the inch within which fracture occurred. This procedure is necessary in order 
to nullify the variations due to the position of the fracture, and is employed in most 
important testing laboratories. The measurements were made along the longitudinal 
centre line of each bar, and the breadth of the gap left between the two broken surfaces 
was deducted from the total length. The gap is usually broadest at about the middle 
of the width of the bar, owing to the fact that fracture generally begins there, and that 
the neighbouring unbroken parts continue to stretch until they are sundered in turn. 
Fig. 4 is an outline of one of the bars after fracture ; the thick irregular line shows the 
gap left when the two broken halves are pressed tightly together. The variation in 
the breadth of this gap with change in the size of test-bar can be obtained from Table 
V. and figs. 8 and 12. 

In Table IV. are given the mean extensions of original lengths of 1, 3, 5, 7, 9, 11, 



202 



MR W. GORDON AND MR G. H. GULLIVER ON 



13, and 15 inches for each series of bars. The extension on a 1-inch length is the 
measured extension of the inch within which fracture took place ; that on a 3-inch 
length is the extension of the fractured inch plus the extension of the inch on each 
side ; the extension on a 5-inch length is the extension on the 3-inch length plus 
the extension of the next inch at each end, and so on. In the case where, owing to 




Fig. 4. — Outline of flat test-piece, originally 3| inches wide and 
J inch thick, showing the gap left between the fractured 
surfaces. The variation of thickness in the constricted region 
is indicated on the contour lines in hundredths of an inch. 



the fracture being nearer one end than the other, a datum inch is available only on one 
side of the previously measured length, twice the extension of this inch is added to the 
previous total. The extension in all cases is expressed as a percentage of the original 
gauge-length. The extension of the first inch is subject to error since the variation of 
position of the fracture within this length has been neglected ; this error is more 
important in the wider bars. In the extension of longer lengths the error diminishes 
rapidly, and is negligible for greater lengths than 5 inches, even in the widest bars 
used ; it is therefore of little consequence from the present standpoint. Another cause 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



203 




m 



n 



CM 



O 



3 

60 

g 

^3 



10 



-° rfl 







S 1 




(f) 


CD 


Ul 

o 


■+J> CO 

° a 

-2 s 




z 


« r2 


CO 


1 


I 8> 




I 


" 'a 




H 


S 3 


K 


O 






z 


60 .5 




UJ 


o» H* 




-1 


1 ' ^-t 


CD 




,fH in 

at CO 

C8 & 



CO 



CM 



o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


CD 


00 


N 


CO 


in 


<t 


CO 


M 



I 

id 



1N33 H3d N0ISN31X3 



TRANS. ROY. S0O EDIN., VOL. XLVIII. PART I. (NO. 10). 



32 



204 



MR W. GORDON AND MR G. H. GULLIVER ON 



TABLE IV. 

Extension op Bars of Different Widths, Measured upon Various Fixed Lengths. 











Extension 


per cent 


. on a L 


angfch of 






Nominal Width. 
Tnches. 


Width 
Thickness' 











































1 


3 


5 


7 


9 


11 


13 


15 Inches. 


§ 


1-76 


179 


45-0 


30-2 


25-3 


23-8 


22 - 2 


209 


19-9 


19-1 


1 


3-81 


147 


50-0 


34 2 


28-8 


26 


1 


24-7 


23 


8 


22-8 


222 


U 


5-64 


128 


55-0 


38-9 


33-0 


29 


4 


27-1 


25 


5 


24-3 


23-3 


2 


7-57 


122 


59-7 


41-7 


34-7 


31 





28-2 


26 


5 


25-2 


24-0 


H 


9-57 


128 


65-0 


432 


35-3 


31 


4 


28-6 


26 


8 


253 


24-2 


3 


11-14 


117 


62-3 


42-8 


351 


31 


2 


28-2 


26 


2 


24-6 


23-4 


3| 


13-32 


127 


67-0 


44-8 


36-5 


32 


4 


29-3 


27 





25-3 


24-0 


4 


15-39 


121 


68-0 


47 3 


38-6 


344 


3L2 


29-1 


27-1 


25-7 



of irregularity is the presence of one or more incipient constrictions along the length 
of a bar, other than that at which rupture has ultimately developed. Such a constric- 
tion gives rise to a local increase in the extension, and to a corresponding peak in the 
curve, but this is smoothed out by taking the average for each set of three bars. No 
allowance has been made for these small constrictions since they cannot be prevented 

I90r 



i- 
z 

UJ 

o 

a 

0. 

z 
o 

co 

z 

UJ 
h 
X 

UJ 



150 



10 



70. 


































o 















w 





















8 10 

WIDTH 
THICKNELSS 



12 



14 



16 



18 



Fig. 6. — Variation of maximum extension at fracture of Hat steel test-pieces with change in the ratio of width to thickness 

from 2 to 16, and a constant thickness of £ inch. 

in an ordinary test. The differences between the extensions of the bars in each set are 
due chiefly to these two causes. 

In fig. 5 are plotted the figures of Table IV. for each width of bar. The extension 
for all bars decreases at first very rapidly from the fracture, and then more slowly, 
tending towards a minimum which is probably about constant for all. Generally 
speaking, the curve of extension for one bar lies completely above that for another of 
less width, but this is reversed in the case of the 2^- and 3-inch bars. 

The extension at zero length, i.e. at the position of fracture, has been deduced from 
the area of the minimum cross-section. Let <J\ be the increase in length of an original 
length <U, through which the fracture passes. Let O be the original area of cross- 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



205 



section, and Q' the final area of cross-section of the bar. Then, if the volume of the 
metal remains constant, a condition which is nearly fulfilled, — 

SI. a = (81+ 8k) . O' 
or the extension is, 

8k _ n - o' 
8/ ~ flT~ ' 

This gives approximately the maximum limiting extension. The mean values are 



70 



60 



50 

H 
Z. 
Id 

a 

£5 40 
a 



z 
o 
01 

z 



30 



20- 



10 



V--- 




L=l inch 



L=3inches 

L=5 inches 

L=7 inches 

L=9 inches 
L=ll inches 
L=l3inches 
L=l5inches 



o 



i 



8 
WIDTH 



10 



12 



14 



THICKNESS 

Fig. 7. — Variation of extension of flat steel test-pieces with change in the ratio of width to thickness from 2 to 16, a 
constant thickness of \ inch, and gauge-lengths of 1 inch to 15 inches. 

given in Table VIII. and are plotted in fig. 6 ; there is a sharp drop at first, but lor 
the bars wider than 1 inch the values are nearly constant. 

Fig. 7 shows the data of Table IV. plotted, not for each bar separately, but for each 
fixed length of bar, and it illustrates the manner in which the extension of a fixed 
length varies as the width of the bar is increased. The curves, excluding that for a 
1-inch length, are all of the same type ; the extension rises at first as the width of the 
bar is increased, then remains almost constant, and finally rises again at about the 
same rate as at first. In fig. 8 is drawn separately, with a more open vertical scale, the 
curve of extension for a length of 8 inches, the standard fixed length for flat test-pieces 
m this country. The vertical distance between the two curves of fig. 8 represents the 



20G 



MR W. GORDON AND MR G. H. GULLIVER ON 



breadth of the gap at fracture, expressed as a percentage of 8 inches. The numerical 



data are given in Table V. 



36 



34 



32 

i- 
z 
ui 
O 

o: 
id 

a 30 



■z. 
u 

x 2 

UJ 

< 

o 



26 



24 



22 



> — 










. 




































o ^- — " 












v 


'V 
















^ 
p 
































// 























8 10 

WIDTH 



12 



14 



16 



18 



THICKNESS 
Fig. 8.— Similar to fig. 7, but with a gauge-length of 8 inches. The vertical distance between the two curves is the breadth 

of the gap at fracture, expressed as a percentage of 8 inches. 

TABLE V. 

Extension of Bars of Various Widths, Measured upon Lengths of 8 Inches, and 11-3VArea. 



Nominal Widtli. 
Indies. 

\ 

1 

1> 
2" 

H 

3 

H 

4 


AVidth 
Thickness' 


Extension per cent. 


Gauge-length = 8 Inches. 


Gauge-length = 11 3 \/Area. 


Gap included. 


Gap deducted. 


Gap included. 


Gap deducted. 


1-76 

3-81 

5 64 
7-57 

9-57 
11-14 
13-32 

15-:;!) 


22-58 
26-15 
28-98 
MC-44 
30-61 
31-10 
3 1 -90 
33-93 


22-29 
25-69 
28-15 
29-48 
29-61 
29-35 
30-71 
32-47 


28-18 
28-79 
30-21 
30-44 
29-39 
28-75 
28-17 
2970 


27-57 
28-13 
29-24 
29-4S 
28-51 
27-32 
27-28 
28-68 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



207 



The form of the curves of extension given in fisfs. 7 and 8 differs from that obtained 
by Barba (fig. 1), in showing no well-marked maximum. If the process of deformation 
were the same in all the bars, so that the geometrical shape of the constricted region 
remained invariable, the extension of a fixed length would increase continuously with 
increase in the width of the bar. The fact that it does not indicates that the shape of 
the constricted region varies in such a way that there is less constriction, and therefore 
less corresponding extension, in a wide bar than in a narrow one. But, on the other 



70 



60 



50 

Z 
bl 
U 

a. 

u 40 

Q. 



Z. 

o 

V) 

^30 



20- 



10 








8 
WIDTH 



10 



12 



14 



16 



THICKNESS 

Fig. 9. — Variation of extension of fiat steel test-pieces with change in the ratio of width to thickness from 2 to 16, a constant 

thickness of J inch, and gauge-lengths proportional to the ratio width/thickness expressed in inches. 

hand, the extension of a fixed length of a wide bar is greater than that of a narrow bar 
because the length is relatively less. These two factors which influence the extension 
act therefore in opposite directions, and from the shape of the curves of figs. 7 and 8 
it is evident that these factors approximately equalise each other over a certain range 
of the ratio width/thickness, namely from about 7 to 12. 

That there is less constriction of the wider bars may be shown in two ways. In 
fig. 9 are plotted curves, from the values given in Table VI., showing the extension of 
lengths proportional to the widths of the bars ; the extension diminishes generally as 



208 



MR W. GORDON AND MR G. H. GULLIVER ON 



the width increases, though in a somewhat irregular manner. The more striking method 
is to compare the actual shapes of the fractured bars. This has been done in 
fig. 10, where the mean outlines of the bars originally about ^, 1,2, and 4 inches 
wide are drawn to the scale of their original dimensions, all reduced to the same size 
for convenience of comparison. It is evident that the proportional reduction in width 



1 u 

0-9 








0-8 








07 


\^~>*~^ ^S^ 

"1^^ 


10 










JO-6 








Q 


- 


09 




5 








-J 0-5 


jJ^s^- : ~ :: ~ ::= ^-^~ — — ^-~^~~~ 


0-8 




< 


^*0" J **' — ^- ^~ — J -"-~""' — ' 






z 






CO 




CD 


/ S \" jr ^^"^ 


°- 7 c/) 


I/) 
UJ 




DC 


/ / y"^t ^^^ 


co 


z: 




O0-4 


'// /J^^ 


hi 

0'6z 
o 


o 

X 








05 i 


H 




03 




1- 


_l 






- 


0-4^ 


< 








2 


CD 


* 


0-2 




0-3 U - 


DC 

O 






- 


0-2 




01 










i i i i i 


01 

i i 


I i 



4" 

2''and\' 



CENTRE 

OF 

FRACTURE 



01 0-2 0-3 0-4 0-5 06 

LENGTH 



0-7 



08 



0-9 



1-0 



WIDTH 

Fig. 10. — Mean profiles of width of flat steel test-pieces after fracture ; measurements taken from bars having widths of 

J, 1, 2, and 4 inches, and a constant thickness of 4 inch. 

Inset. — Mean profiles of thickness of same bars, measured on longitudinal centre line. 

is less the wider the bar, and that there is consequently less extension of the wider 
bars in the constricted region. The inset in fig. 10 shows in a similar manner the mean 
thickness of the same four sets of bars, as measured along the longitudinal centre line ; 
the order of the curves is just the same as for the width. It is interesting to note that 
tlie final minimum thickness of all the bars measured at the broken surface is nearly 
constant. For the larger diagram the origin is at the centre of the gap, while for the 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



209 



inset the origin is shifted slightly so as to correspond with the edge of the broken 
surface. In both cases the cross dimensions have been set up from a straight base, so 
that the figures do not represent the true shape of the bars. The profiles of the 1^-, 2-^-, 
3-, and 3^-inch bars have been omitted for the sake of clearness. 



TABLE VI. 

Extension of Bars of Different Widths, Measured upon Lengths Proportional 

Width 



to the Ratio 



Thickness' 







Extension per cent, on a Length equal to 


Nominal Width. 
Inches. 


Width 
Thickness' 








Width 
Thickness' 


Width 
2 Thickness' 


, Width . , 
i"m, inches. 
Thickness 


\ 


1-76 


37-0 


46-5 


56-8 


1 


3-81 


31-3 


40-5 


50-6 


1* 


564 


316 


39-9 


49-9 


2 


757 


30-2 


38-4 


49 


H 


9-57 


28-1 


35-8 


47-5 


3 


11-14 


261 


33-6 


437 


n 


13-32 


24-9 


325 


42-8 


4 


15-39 


25-6 


32-9 


42-4 



In fig. 4 the actual outline of one of the bars is shown, and the variations in thick- 
ness are indicated by the contour lines, drawn at intervals of "02 inch (the contour for 
a thickness of 0'21 inch is given also). There are evidently two nearly straight 
depressions crossing the bar, and intersecting each other at about the middle of its 
width. This peculiar phenomenon, which is a characteristic of flat metal bars broken 
by tension, is known as the Contractile Cross ; its essential features have been described 
elsewhere (5). Fracture takes place frequently, but not invariably, along one of the 
grooves ; in the figure the line of fracture has followed one groove, and the bottom of 
the other groove is indicated by the oblique dotted line. 

The extensions, as measured on various lengths proportional to the square root of 
the cross-sectional area of each bar, are given in Table VII., from which fig. 11 has 
been drawn. The extension is much more nearly constant than in the case of a fixed 
gauge-length, but there are still variations. The change in the form of the constricted 
region has here a greater effect than in the case of the fixed length ; in other words, 
the factor which formerly was sufficient to neutralise the effect of relative decrease of 
length and to keep the extension nearly constant over a certain range, is now able to 
give rise to an actual depression in the value of the extension within that range. The 
curves of extension of fig. 1 1 are therefore of wave form, the extension at first increasing, 
then decreasing, and then increasing again as the ratio width/thickness is increased 
continuously. There is no great difference in the form of the curves for lengths between 
2 ^/area and 1 5 */area. 



210 



MR W. GORDON AND MR G. H. GULLIVER ON 



TABLE VII. 

Extension op Bars of Different Widths, Measured upon Lengths Proportional to VArea. 



Nominal Width. 
Inches. 


Width 
Thickness' 


Extension per cent, on a Length equal to 


vArea. 




5 VArea. 


10 ^Area. 


15 VArea. 


2 VArea. 


h 

1 

H 

2 

H 

3 
31 

4 


1-76 

3-81 

5-64 

7-57 

957 

11-14 

13-32 

15-39 


59-0 
59-2 
61-6 
650 
68-8 
64-7 
67-8 
67-6 


50-0 
500 
52-2 
53-8 
55-7 
52-2 
54-0 
55-0 


37-7 
36-6 
38-7 
39-3 
38-4 
36-8 
37-2 
38-4 


28-7 
28-8 
30-7 
30-8 
29-8 
28-6 
28-6 
29-8 


25-3 

25-7 
26-8 
26-8 
26-0 
24-6 
24-3 
25-6 



70 



60 



. 50 

i- 
z 
ul 
o 

DC 
£ 40 

z 
o 
(f) 

£ 30 



20 



10 





I 



8 
WIDTH 



10 



12 



14 



THICKNESS 

PlO. 11. — Variation of extension of flat steel test-pieces with change in the ratio of width to thickness from '2 to 16, a 
constant thickness of 4 inch, and gauge-lengths of \Zarea to 15\/area. 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



211 



In fig. 12 the curve of extension for a length equal to 11'3^/area, the Continental 
standard variable length, is plotted separately with a more open vertical scale ; it has 

32 



o 

w 

z 
u 

I- 

X 

u 

_l 

< 

H 
O 



30 



28 



26 



24 



22 







c/X 
















^ 




? 














\^l*> 


Ji v 























































8 10 

WIDTH 



12 



14 



18 



THICKNESS 
Fig. 12. — Similar to fig. 11, but with a gauge-length of 11 '3 \Jaxea. The vertical distance between the two curves is the 
breadth of the gap at fracture, expressed as a percentage of ll'3V*rea. 

much the same characteristics as those of fig. 11. Notice may be taken of the fact that 
the breadth of the gap varies almost directly as ^/area, the two curves of fig. 12 being 
nearly parallel. The data for fig. 12 are given in Table V. 



70 



60 



z 50 

o 

t- 
u 

3 

a 

id 



40 



O o 

o o 

o 



4 



8 10 

WIDTH 



12 



14 



16 



18 



THICKNESS 

Fig. 13. — Variation in reduction of area of flat steel test-pieces, with change in the ratio of width to thickness from 

2 to 16, and a constant thickness of \ inch. 

6. Keduction of Area. 
The reduction of area is the difference between the original cross-sectional area 
of the bar and the minimum area measured after fracture, expressed as a fraction 

Q.-Q! 



of the original area ; that is, the reduction of area is 2 = 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 10). 



o 



, where O and O' 

33 



212 



MR W. GORDON AND MR G. H. GULLIVER ON 



represent the original and the final area, respectively. The values of this quantity- 
are given in Table VIII., and a curve showing its variation with the ratio width/ 
thickness is drawn in fig. 13. The curve is of the same form as that of fig. 6, the 
curve of maximum extension at fracture, the two quantities being related thus : — 



8A. 2 
Maximum extension, — = 



TABLE VIII. 
Reduction of Area, and Maximum Extension at Fracture. 



Nominal Width. 
Inches. 


Width 
Thickness 


Reduction of Area. 
Per cent. 


Maximum Extension 

at Fracture. 

Per cent. 


l 

2 


1-76 


641 


178-7 


1 


381 


59-5 


147-8 


1* 


5-64 


56 2 


127 9 


2 


7-57 


54-9 


121-5 


H 


9-57 


56-2 


128-8 


3 


11-14 


54-0 


1171 


H 


13-32 . 


56-0 


127-3 


4 


15-39 


54-7 


120-8 



The method of determining the final area has been described already, and the diffi- 
culty of securing accuracy has been mentioned. In the neighbourhood of fracture the 
section of a bar initially rectangular is not a rectangle, but has curved sides, the curva- 
ture being less pronounced the wider the bar. Moreover, the ratio of width/thickness does 



< 



140 












n 











120 






o^ 




, ■ ' "o 


:v)RE__^- 


o 












* 
1 
1 






OUTSIDE 


0F CONSTP 


ICTION 








100 


i 


>' 















I 



8 



WIDTH 



10 



12 



14 



16 



18 



Tig. 14. 



THICKNESS 

-Change in the ratio of width to thickness of flat steel test-pieces after fracture. The upper curve shows the increase 
in the ratio at the fracture itself, and the lower one the increase in the sensibly parallel portion of the bar. 



not remain constant but increases during extension ; in other words, the metal is reduced 
relatively more in thickness than in width, a result quite to be expected. Similarly the 
width is reduced less in the wide than in the narrow bars, even when compared with 
the diminished thickness. The values of the final ratio of width to thickness are given 



THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 



213 



in Table IX., both at the fracture and at a point originally distant therefrom a length 
equal to the width of the bar. From fig. 14, in which these values are plotted, it is 
seen that the change in the ratio increases fairly continuously with the ratio itself as 
regards the measurements at fracture, while in the sensibly straight portion of the bar 
the increase of the ratio is almost constant at about 3^ per cent. 



TABLE IX. 

Initial and Final Ratios of Width to Thickness. 





Width/Thickness. 














Final ratio / Initial ratio. 
Per cent. 








Nominal Width. 




Final. 




Inches. 


Initial. 




















At Fracture. 


Outside of 
Constriction. 


At Fracture. 


Outside of 
Constriction. 


l 

2 


1-76 


2-07 


1-82 


1176 


103-4 


1 


3-81 


4-85 


3-95 


127-3 


103-7 


H 


5-64 


730 


5-85 


129-4 


103-7 


2 


7-57 


9-91 


7-86 


130-9 


103-8 


H 


9-57 


13-53 


9-88 


141-4 


103-2 


3 


11-14 


15-01 


11-50 


134-8 


103-2 


H 


13 32 


19-50 


1385 


146-4 


1040 


4 


15-39 


21-60 


15-92 


140-4 


103-4 



Summary. 

In a series of flat test-bars of mild steel, of constant thickness and variable width, 
the effect of the ratio of width to thickness upon the apparent strength and ductility 
of the metal has been determined. 

(1) The strength, both elastic and ultimate, is sensibly unaffected. 

(2) The extension, measured on a fixed gauge-length of 8 inches, increases as the 
ratio width/thickness is increased from 2 to 7, remains sensibly constant from 7 to 12, 
and rises again as the ratio passes from 12 to 16. The extreme difference of extension 
is 10 per cent, on 8 inches, that is, nearly one-half of the extension of the narrowest bar. 

The extension on other fixed gauge-lengths varies in a manner closely similar to 
the above. 

(3) The extension, measured on a variable gauge-length equal to some definite 
proportion of the ratio width/thickness expressed in inches, decreases somewhat irregu- 
larly as the ratio itself increases. 

(4) The extension, measured on a variable gauge-length equal to 11'3 */area, 
increases as the ratio width /thickness is increased from 2 to about 7, then decreases 
from 7 to about 12, and then increases as the ratio passes from 12 to 16. The extreme 



214 THE STRENGTH AND DUCTILITY OF FLAT STEEL BARS. 

difference of extension is only 2 per cent, on the gauge-length, that is, about one- 
fourteenth of the extension of the narrowest bar. 

The extension on other variable gauge-lengths proportional to Varea changes in a 
somewhat similar manner. 

(5) The reduction of area decreases as the ratio width/thickness is increased from 
2 to 6, and then remains sensibly constant. 

(6) The thickness of a flat bar is reduced relatively more than the width, and the 
width of a narrow bar is reduced relatively more than that of a wide one. 

(7) The critical ratios of width to thickness are not strongly marked, and probably 
vary with the absolute thickness of the bar, and with the quality of the metal. 

In conclusion, the thanks of the authors are due to Prof. Hudson Beare for the 
facilities accorded, and for the interest which he has shown in the work. 



REFERENCES. 

(1) Publications of the Engineering Standards Committee, No. 18, Crosby Lockwood, London, 

June 1907. 

(2) Barba, Commission des Methodes d'Essai des Materiaux de Construction, Rothschild, Paris, 1895. 

III., A, 5. 

(3) Appleby, Proc. Inst. Civ. Eng., 1894, cxviii., 395. 

(4) Unwin, Proc. Inst. Civ. Eng., 1903, civ., 170. 

(5) Gulliver, Proc. Inst. Mech. Eng., 1905, 141, and 1907, 519. 



( 215 ) 



XI.— A Monograph on the general Morphology of the Myxinoid Fishes, based on 
a study of Myxine. Part IV. — On some Peculiarities of the Afferent and 
Efferent Branchial Arteries of Myxine. By F. J. Cole, D.Sc. Oxon., Professor 
of Zoology, University College, Reading. Communicated by Dr R. H. Traquair, 
F.R.S. (With One Plate.) 

(MS. received November 21, 1911. Read January 8, 1912. Issued separately April 1, 1912.) 

The first three parts of this work, on the skeleton and muscles, were published in 
the Transactions of the Society in 1905, 1907, and 1909. 

In 1905* I briefly drew attention to the existence of vascular papillae on the 
afferent branchial arteries of Myxine, which had up to that time escaped notice, and 
which I then described in the following words : " In addition to the definite blood- 
vessels, Myxine possesses a system of large lacunar spaces, such as the extensive sub- 
dermal cavity, the spongy tissue of the head, the peribranchial sinuses, etc., which have 
been generally regarded as belonging to the lymphatic system. I have, however, long 
been convinced that these spaces were in communication with the blood vascular stream. 
Jackson, in his work on the vascular system of Bdellostoma, mentions the passage of 
injection mass from the vessels into the lymphatics, but believes the connection between 
the two to have been an artificial one, since he does not find red blood corpuscles in 
the lymphatics in fresh and uninjected material. If this is true, then Myxine is greatly 
different, as blood is invariably to be found in the lymphatics in living material. 
Ewart, in his paper on the vascular peribranchial spaces in the Lamprey, correctly 
appreciates the situation, and explains the general appearance of blood in these spaces 
by connections between them and the internal jugular vein found by him. 

" I shall enter fully into the morphology of these spaces in my fourth part on the 
vascular system of Myxine. In the meantime I may mention that I have discovered 
projecting from the posterior surface of each afferent branchial artery, at the place 
where this artery enters the gill sac, one or more papillae, and I have found these in 
every specimen which has been dissected for them. On cutting serial sections of several 
of these papillae it is seen that the base of each is widely excavated, and is, in fact, an 
evaginated portion of the cavity of the artery, whilst from this excavation there passes 
to open on to the exterior one or more fine channels lined by epithelium. The calibre 
of these channels is usually only slightly in excess of the width of an average red blood 
corpuscle. The presence of these channels at once explains the appearance of blood in 
the peribranchial sinuses in the normal living fish, and it seems certain that there must 
be other connections between the blood-vessels and the so-called lymphatic spaces in 
other parts of the body." 

* Anat. Anz., Bd. xxvii. pp. 325-6. 
TRANS. ROY. SOC. EDIN., VOL. XLVIIL PART I. (NO. 11). 34 



216 PROFESSOR FRANK J. COLE 

Since writing the above I have examined the curious structures in question much more 
closely by means of serial sections, injections of various fluids and masses, and by whole 
preparations, and I am now able to add considerably to the description just quoted. 

1. Methods. 

It will readily be admitted that if the vascular papillae are in open communication 
with the surrounding spaces, so that blood may pass from the arteries into the lymph 
sinuses or pleural sacs by which the gills are enclosed, we have to do with a fact so 
unexpected that nothing short of complete demonstration will ensure its acceptance. 

Assuming such a connection to exist, it may seem at first sight a simple matter to 
establish it. It is, however, by no means easy to do so ; and I had continuously investi- 
gated these structures for some time, and had actually written withdrawing my pre- 
liminary statement, when some final injections settled the matter in favour of its 
accuracy. Such a contradictory result is partly accounted for by the fact that whilst 
in some papillae there is an undoubted communication between the artery and the 
sinus, in others, and perhaps the majority, the communication has been closed, and 
the papillae are vestigial structures. 

My preliminary statement was based on the study of serial sections ; and if in some 
of those sections (e.g. figs. 2-5) the existence of the communication seems beyond 
question, it is so easy to misinterpret the sections that such evidence in itself cannot 
be held to be convincing. 

Injections of preserved material are useful, but not decisive. I have tried a large 
number of such injections with all the ordinary injection fluids and masses, and have 
never once found the medium to pass right through the gills into the efferent arteries. 
Hence negative results obtained by this method prove nothing either way. An 
injection, to be completely successful, must be carried out on an animal as soon as 
possible after it has been removed from the sea, and immediately after death, with the 
heart still beating. One can, in fact, almost deduce the time the animal has been dead 
by the speed with which it may be injected. A syringe injection may be so rapid that 
some of my first attempts, put on one side as failures owing to apparent leaking, were 
afterwards found to be injected throughout the whole body, and the leakage due to the 
injection mass, having returned to the heart, escaping through the cut auricle. 

Myxine are not only difficult to kill — like so many marine animals in captivity they 
will neither live nor die — but their reflexes remain functional for a long time. This spoils 
many an injection. After trying a number of methods, I find the best is to snip off 
the tail immediately behind the cloaca, and to immerse the animal in warm sea-water. 
The latter to a certain extent acts as an anaesthetic, whilst the main blood-vessels are 
drained, and the tail can be ligatured as soon as the injection mass begins to escape. 

Gelatine masses are unsatisfactory for two reasons : (1) they must be thrown in 
hot, and the passage of a hot mass through the vessels of a cold-blooded animal affects 
the elasticity of their walls so as to almost vitiate the result where very fine channels 



ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 217 

are concerned ; (2) subsequent treatment is difficult. The mass expands and contracts 
according to the amount of water present. It contracts considerably on dehydration, 
and expands again if the sections are stretched on water on the slide. Preparations are 
completely ruined as regards the finer details if they are placed after injection in formalin 
or weak alcohol, owing to the excess of water producing expansion of the gelatine and 
consequent rupture of the tissues. Further, gelatine becomes brittle and difficult to 
handle in paraffin. 

Syringes must be avoided. It is difficult to regulate the pressure, and still more 
difficult to inject with the minimum pressure required. Syringe injections generally 
"blow off" the tops of the papillae, and what appears to be a demonstration of the 
communication is only an illustration of the crudeness of the method. 

What I regard as a convincing proof of a connection between the artery and the 
surrounding sinus via the papillae may be readily obtained in the following way. The 
injection fluid is a simple solution of Prussian blue in distilled water. This is placed 
in a glass vessel which is raised about five feet above the laboratory bench, and an 
india-rubber tube leads from it to the bench, and has at its end a glass cannula drawn 
out to a capillary tube. The flexible tubing is closed by a screw clamp placed just 
above the cannula. The base of the latter may be passed through a cork, so that when 
it is in position in the vessel from which the injection is to be made, the whole can be 
held and kept steady by the clamp of a retort stand, leaving both hands of the observer 
free. Immediately the animal is motionless it is strapped down under sea-water, and 
the ventricle and ventral aorta are exposed. The free extremity of the ventricle is 
snipped off, and the cannula pushed into the root of the aorta and. a ligature applied. 
The afferent branchial arteries may now be carefully exposed under a Braus-Druner 
dissecting microscope by removing the external walls of the pleural sacs, and the whole 
of the subsequent operations kept under continuous microscopic observation, since all 
that is needed to start the injection is to slightly turn the screw of the clamp. The 
speed of the injection is thus under complete control, but to demonstrate the openings 
of the papillae without rupture, as the latter are clearly the weakest points of the arteries, 
the injection fluid should be allowed to pass very slowly into the ventral aorta. Watch- 
ing the papillae under the microscope, and in the living animal they are more or less 
transparent, the injection is seen to pass immediately into the large cavity at the base 
of the papillae (figs. 2, 3, 5). Then fine blue threads spring out, connecting this cavity 
with the apex of the papilla. There is a slight pause, and finally delicate spirts of 
blue are discharged from the papilla into the cavity of the surrounding sinus. It seems, 
therefore, that when the pressure in the artery is low, there will be no discharge of 
blood into the sinus, but that when the pressure rises beyond a certain limit, blood is 
transferred from the cavity of the artery to that of the peri-branchial lymph sinus or 
pleural sac. 

The use of an injection medium which is a cold solution, ensures an easy passage into 
the finest vessels, and as it is readily precipitated by the addition of alcohol — in fact it may 



218 PROFESSOR FRANK J. COLE 

be watched precipitating as it travels through the vessels, — serial sections can be cut of 
the tissues ; and as far as the finer vessels are concerned, this injection is as satisfactory 
as any. An additional advantage is that it exercises no staining effect on the tissues, nor 
is it itself stained by histological dyes. The walls of the vessels can therefore be stained 
red, and preparations obtained which are both clear and convincing. The only objection 
to its use — and it is unfortunately a serious one — is its tendency to precipitate during 
the operation by combining with the organic fluids with which it is brought into contact, 
and therefore to block in the smaller vessels. This may be avoided by thoroughly 
washing out the vessels before throwing in the blue — an important precaution, since the 
precipitate is of a clinging, flocculent, obstructive nature. I find it convenient to use a 
cannula with a double lead. Two different fluids can then be employed successively 
without unshipping the cannula. 

2. General Anatomy of the Parts. 

The ventral or cardiac aorta of Myxine (fig. 1, c. ao.) courses directly forward from 
the ventricle of the heart {vent.), giving off on each side as it goes along the afferent 
branchial arteries (af. br.). The walls of the latter have capillaries, and are in fact 
richly vascular. Each afferent vessel passes straight up to its gill, and enters it below 
and slightly behind the exit of the efferent gill duct. 

Examination of a large number of specimens reveals peculiarities in the first and 
last afferent branchial arteries * which are of very common, even if not of universal, 
occurrence. The former, on its way up to the gill, gives off a very short anterior twig 
(fig. 1, # ), which varies somewhat in its structure, but is always blind and connected 
with the neighbouring tissues by one or more fine but strong threads. We know from 
Stockard's work that a hyomandibular and two post-hyomandibular clefts develop and 
then degenerate between the mouth and the first functional gill of Bdellostoma, and it 
is therefore possible that this twig may represent the vestiges of the arterial supply of 
these structures. An additional reason for this view is found in the variations of the 
twig itself. Sometimes there is more than one arising from the arterial trunk, and when 
that occurs they soon meet and fuse. Or the twig may give off some branches large 
enough to be easily recognisable. Or, most important of all, it may terminate in a 
capillary network or rete mirable, which thus represents all that is left of the gill. 
This network, when present, is situated in a pear-shaped expansion at the extremity of 
the twig, and the network itself is easily demonstrable by injection. I have never found 
any traces of an opening to this twig, and we note that it is situated further down on 
the arch than the vascular papillse. I therefore consider it to represent a different type 
of structure, although it may not be homologous in all specimens. 

J. MiJLLER, who noticed this structure both in Bdellostoma and Myxine,^ finds in 
the Cape Bdellostoma a connection between it and the carotid system by means of a 

* The most anterior or first gill is supplied by the first afferent branchial artery, and so on. 
t Abh. Ah. Berlin, Jahr 1839, p. 191. 



ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 219 

very fine imperforate thread. At its afferent and efferent extremities this connection 
is wider, hollow, and contains blood derived from the corresponding arteries. He there- 
fore regards it, and quite justifiably, as a ductus Botalli or vestigial aortic arch, corre- 
sponding to the one immediately in front of the existing first. In spite of numerous 
and most careful dissections of Myxiiie, I have not found any traces of this connection ; 
nor does Jackson find it in Bdellostoma dombeyi. He says : # "The last [ = my first] 
afferent branchial artery of each side gives off a small branch a short distance from the 
gill. This branch possesses a lumen only at its origin, if at all. It soon becomes reduced 
to a slender string of connective tissue which becomes lost in the connective tissue 
around the ' club muscle.' Attached to this string is a small spheroidal body, apparently 
made up of fibrous and fatty tissue." " In addition to the observation of Muller, I 
have added that a spheroidal or flattened mass of connective tissue is found attached 
to the ' ductus ' a short distance from its origin. This body is larger and more saccular 
in appearance in Bdellostoma forsteri, and evidently may be interpreted as the rudiment 
of the gill pouch corresponding to the obliterated branchial artery." The discovery of 
the rete mirable in Myxine naturally supports Jackson's suggestion, as I have above 
indicated. 

The peculiar feature associated with the last afferent branchial artery occurs 
generally on the left side rather than on the right. The artery usually divides sooner, 
and forms two large vessels, the posterior of which gives off a branch which passes 
backwards and upwards, it may be for quite an appreciable distance ; but sooner or later 
it loses its lumen, becomes thread-like, and finally disappears altogether. Connected 
with it, as a rule, are several vascular papillae (cp. fig. 1, f). The significance of this 
structure becomes obvious when we investigate the variations in the gills. Of the eight 
cases I have carefully examined, six consisted of an extra gill on the left side only, 
and the remaining two of an extra gill on both sides. Also, in five out of these eight 
cases the extra gill was supplied by a branch from the last afferent branchial. Again, 
in another case, with the normal number of gills, there were only five afferent branchial 
arteries, the sixth gill being supplied by a branch from the last of these. I think, there- 
fore, that the blind twig from the last afferent branchial artery is associated with the 
former existence of a seventh pair of gills. 

The structures which I have called vascular papillae were first described in my pre- 
liminary paper. They are found both on the afferent and efferent branchial system 
(fig. 1), but are larger and more conspicuous on the former. They vary somewhat con- 
siderably both in number and structure, but I have never found them absent on a single 
occasion, although a very large number of individuals have been examined. The reason 
why they have hitherto been overlooked on the afferent vessels (which are more usually 
dissected) is due, doubtless, to their position. As a rule, they are situated high up on 
the artery, near the point where it disappears into the gill, and are therefore tucked 
away under the efferent gill duct (e. g. d.) ; but they do on occasion occur lower down, 

* Univ. Cincinnati Bull., No. 5, 1901, pp. 21, 37. 



220 PROFESSOR FRANK J. COLE 

as shown in the second afferent branchial of fig. 1. Generally, also, they are found on 
the posterior surface of the artery, and are hence directed backwards. Their structure, 
which will be described in detail later, varies from a simple papillae with an unbranched 
cavity to an elaborate digitiform structure with a complex cavity. 

The efferent branchial arteries of Myxine (fig. 1, ef. br.) differ from the afi'erents 
in so far as there are two of them to each gill. In Bdellostoma dombeyi there is only 
one to each gill pouch, but there are two in B. for stem. In Myxine each afferent gill 
duct (a. g. d.) has an artery immediately in front of and behind it {ef br.). All the 
efferent arteries open into a commissural vessel known as the common carotid (c. car.), 
which is attached to the side of the oesophagus (oes.), in front, passing straight forwards 
to the head, whilst behind, it rises to open into the systemic aorta (s. ao.) just posterior 
to the sixth gill. The systemic aorta is prolonged forwards in the median line over 
the gut as the anterior systemic aorta (a. s. ao.). 

The common carotid and anterior systemic aorta are connected up by three anas- 
tomoses, and although I have found this number to be very constant, the same three 
are not always present. For example, in fig. 1 the anastomoses are connected with 
afferent gill ducts three, four, and five, but they may be associated with four, five, and 
six. It seems, on the whole, probable that each anastomosis is formed either by the 
fusion of a pair of efferent arteries, or represents the dorsal extension of one. 

It is hardly likely that the common carotid has any existence per se, but stands for 
merely a series of longitudinal anastomoses. This is borne out by the state of affairs 
both in Myxine and Bdellostoma. In the former the common carotid usually narrows 
down behind the fourth afferent gill duct (cp. fig. 1), and between the fifth and sixth 
gills it may become so fine a thread as to be almost imperforate, in such a case the 
last two transverse anastomoses are clearly the direct continuations of efferent branchial 
arteries, and the posterior section of the carotid which rises to fuse with the aorta is 
obviously the continuation of the last efferent branchial. 

There are no vascular papillae on any part of the dorsal aorta, either in the branchial 
region or anteriorly to it. Dorsally the papillae are confined to the common carotids 
(fig. 1), but are by no means restricted to the branchial region. The common carotid 
between the first gill and the division far forwards into external and internal carotids, 
always bears a number of the papillae. They are, however, largely confined to the posterior 
four-fifths of this anterior section of the carotid, as the following count, from behind 
forwards, exemplifies : 8, 11, 7, 4, 2. The pre-branchial papillae are simpler in structure 
than those occurring in the immediate neighbourhood of the gills, and if they open 
externally at all, it must be very rarely. 

It now becomes necessary to describe the spaces into which blood is discharged 
by the afferent and efferent branchial arteries through the agency of the vascular 
papillae. In front of the branchial region, i.e. in the region of the club muscle, there is 
a large dorso-ventrally flattened lymph sac between the notochord and the oesophagus, 
but which extends some distance laterally on each side of the gut. The anterior 



ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 221 

median aorta and the paired common carotids lie morphologically outside the roof and 
floor of this sinus respectively. Below the gut is another large sinus of similar form, 
which covers practically the dorsal surface of the club-shaped muscle.* Laterally 
to the gut these two sacs are separated by a thick zone of fatty tissue, but, as it is 
easy to demonstrate by inflation and by injection, they are in frequent and wide 
communication by dorso-ventral channels situated at the lateral margins of the 
oesophagus. 

Behind the posterior extremity of the club muscle, in the region of the gills, the 
ventral of these sacs becomes broken up and finally disappears, but the dorsal sinus is 
prolonged backwards unmodified throughout the gill region, and acquires connections 
with the peribranchial sacs by vertical channels which accompany the efferent branchial 
arteries. The contents of the latter sac, therefore, may pass quite freely into the dorsal 
oesophageal sinus. 

The peribranchial or pleural sacs or sinuses f are large and well-defined bags 
covered by a fatty tissue — one to each gill. They are invariably distended with 
injection mass when the vessels have been filled from the ventricle of the heart, owing 
to the mass passing through the vascular papillae into the sacs. The gill projects quite 
freely into the sac, the walls of the two structures being here and there connected by 
fine tough threads. The posterior wall of one sac is very closely opposed to the anterior 
wall of the one behind, so that there is a double partition between any two adjacent 
gills. This partition is quite imperforate, so that no communication is here possible 
between contiguous sacs. 

There can be no doubt, as Johannes Muller first pointed out, that the gills are not 
morphologically within the pleural sacs, but that they are related to the sacs in 
precisely the same manner as the abdominal and thoracic viscera are to their respective 
cavities. In other words, each pleural sac consists of a visceral and a parietal layer, and 
the .gill, with its afferent and efferent ducts and arteries, is really outside it. This can 
be easily demonstrated by serial sections. J. Muller, in fact, compares the peribranchial 
sinuses of Cyclostomes not with lymph sacs properly so called, but with the pleural 
cavities of higher animals. 

The ductus cesophago-cutaneus has no sinus associated with it, but is wedged in 
between the last pleural sac of the left side and the pericardial cavity. 

The first pleural sac extends a little distance in front of the gill it encloses at the 
side of the club muscle, and is overlapped behind by the second pleural sac. it is 
only prolonged a short distance on to its efferent gill duct, the greater part of which, 
therefore, is visible without removing the sinus. 

The second sac spreads over the external surface of the root of its efferent gill duct 
for a short distance, and this tendency for the sac to extend over the efferent duct 
becomes greatly and suddenly emphasised behind the second efferent duct, so that only 
the first two ducts are visible without removing the covering sinus. The sinus system 

* Cp. J. Muller, Abh. Ak. Berlin, 1834, p. 253. t J. Muller, Abh. Ak. Berlin, 1834, p. 264 et seq. 



222 PROFESSOR FRANK J. COLE 

is further complicated by the overlapping of the gills, any gill, except of course the 
first, overlapping externally the one in front as far as the origin of its efferent gill duct. 
The result is that in the posterior gills those parts of the pleural sacs spreading over 
their efferent gill ducts course backwards side by side, and these portions of the sacs 
communicate freely with each other. It is hence not strictly correct to say that the 
gill sacs only communicate indirectly with each other through the medium of the 
longitudinal ventral and dorsal sinuses. 

The last gill sac communicates freely above with the sinus dorsal to the oesophagus, 
which itself ceases to exist behind the branchial region. Posteriorly, the last gill sac 
terminates blindly by a somewhat irregular border. 

In the mid-ventral line, and situated just above the inferior jugular vein, is a large 
longitudinal sinus, in which courses the cardiac aorta. Anteriorly, this sinus splits into 
two wide channels, which pass upwards and forwards in a curve to open into the first- 
pair of pleural sacs. In the neighbourhood of the split another pair of channels are 
given off, which course straight up into the second pair of gill sacs. Similarly, four 
other pairs of channels arise from the median longitudinal sinus to open into the four 
posterior gill sacs. In the main channel and its branches are situated apparently the 
cardiac aorta and the afferent branchial arteries — connected with the walls of the sinus 
by means of numerous fine but strong threads. The exact relation of the arteries to 
the cavities in which they appear to lie can only be ascertained by a study of their 
development, but we can hardly be wrong in concluding, on a 'priori grounds, that the 
arteries are situated morphologically outside the cavities. 

The median sinus is prolonged backwards behind the exit of the last lateral channel 
on to the base of the ventricle, but here its cavity becomes much broken up by numer- 
ous attachments between its visceral and parietal walls. It has no communication 
with the pericardial cavity, nor is there any justification for regarding it as the peri- 
cardium itself. 

In Petromyzon, according to J. Muller, the gills lie in closed sacs, and there is no 
ventral longitudinal sinus. The afferent gill arteries pass to the gills between the two 
abutting walls of contiguous pleural sacs, thus differing from the Myxinoids in a striking 
manner. Although J. Muller, Rathke, and Robin noticed blood in the so-called 
lymph spaces, it was Langerhans # who first established any connection between these 
spaces, which he compared with those of Amphibia, and the blood vascular system. He 
injected the subcutaneous sinus, and found that the injection passed first of all into 
other and internal lymph spaces, and finally reached the veins and heart. EwARi't 
also states, independently of former writers, that the peribranchial spaces of the 
Lamprey normally contain blood. He did not succeed in finding any connection via 
the arteries, but was able to fill the spaces by injection from the veins. He believes 
the veins concerned are the internal jugulars. At about the same time Schneider + 

* Verhand, nat. Gen. Freiburg, Bd. vi., 1873. + Jour. Anal, and Phys., vol. xii., 1877. 

I Beilr. z. vergleich. Anat. u. Entwick. d. IVirbdthiere, Berlin 1879, 



ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 223 

stated that the peribranchial sinuses always contained venous blood, and he investigated 
the connection of the lymph spaces with the veins by the injection of coloured fluid, 
although he states that this connection may easily be followed without any injection. 
He notes, as I have also done in Myxine, that blood may be clearly seen in the 
subcutaneous sinus in the living Petromyzon jluviatilis. 

Recent writers have paid very little attention to the peribranchial spaces of the 
Lamprey. # Gegenbaur, in his text-book,t does no more than quote Johannes Muller. 
Vogt and Yung| state that the sacs are completely closed and contain a viscid liquid, 
probably lymph, which coagulates in spirit into granular yellow masses. I take it the 
yellow colour is due, as in Myxine, to a certain admixture of blood. Vialleton does 
not believe the spaces to belong to the blood vascular system, and returns to the old 
view, as we now know quite erroneously, that the blood they contain has extravasated 
into them through ruptured walls. Cori, § in his important paper on the blood vascular 
system of the young Ammoccetes, only mentions very briefly the system of blood 
sinuses. 

Mozejko has conducted numerous experiments in the injection of the vascular 
system of Petromyzon, || but, apart from the fact that he says it is possible, he does not 
appear to have more than casually used the method of injecting from the heart. Else- 
where 1T he controverts Vialleton's statement that the peribranchial sinuses do not 
normally contain blood, and asserts that red blood corpuscles are found in all the lymph 
cavities. The sinuses cannot be injected directly via the arteries, but are easily filled 
by injection through the veins. The peribranchial sinuses, however, only communicate 
indirectly with the jugular veins, but in the case of some of them there is a communica- 
tion inter se. He regards the lymph sinuses of Petromyzon as blood-vessels and 
lymphatics at the same time. 

In Bdellostoma Jackson remarks : ** " I have observed in several cases a marked 
tendency for the injected carmine gelatine to escape from the blood-vessels into the 
surrounding lymphatics, which are very numerous and extensive. These lymphatic 
spaces, especially the sub-dermal spaces in the caudal region and the peribranchial 
spaces around the gill pouches, are usually found more or less injected, although the 
blood-vessels show no signs of over-distension. The lymphatic spaces around the 
vessels in the gill itself are also often filled. This condition may be interpreted as 
indicating that the capillary walls are unusually weak and permeable, so that the injected 
liquid passes through them, carrying blood corpuscles with it. That this process is not 
normal is shown by the absence of red blood corpuscles from the lymphatic spaces in 
life and in uninjected specimens." In Myxine, as I have indicated above, the presence 
of red blood may easily be seen in the living animal in the subcutaneous sinus, and red 

* Cp. Nestlbr, Arch.f. Naturgesch., 1890 ; Favaro, Atti Accad. set. Veneto-trent.-Istriana, 1905. 
t Bd. ii. p. 221. X Anat. comp. prat., t. ii. p. 460. 

§ Arb. zool. Inst. Wien, t. xvi., 1906. || Z. f. w. Mikrosk., Bd. xxvii. p. 248, 1910. 

1 Anat. Anz., Bd. xxxvi. p. 618, 1910. ** Op. cit., p. 35. 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART I. (NO. 11). 35 



224 PROFESSOR FRANK J. COLE 

blood corpuscles occur in un injected specimens in all the so-called lymphatic spaces of 
the body. It has been stated that the vessels rupture and fill the lymphatics in bring- 
ing the animal up from the bottom. This may conceivably be so at Cullercoats, where 
Myxine are not usually found within the 23-fathom line, and most of my material was 
collected in over 25 fathoms of water ; but the Californian Hag used by Jackson occurs 
quite commonly in from 10 to 20 fathoms of water. If, therefore, the presence of 
injection mass in the lymphatics is due to rupture of the vessels, we are driven to 
distrust any results produced by that method. It will, however, be obvious from the 
preceding discussion that in this matter Jackson is mistaken. 

We have seen that in Myxine blood enters the pleural or peribranchial sacs via 
the vascular papillae on the afferent and efferent arterial system, although chiefly by 
the former. If this be the case, there must be some corresponding connection with 
the venous system by which the contents of the sacs may be drawn off. Some chance 
injections enabled me to demonstrate this connection in a very striking manner. 

Injection by the ventricle of the heart is a little difficult. The heart is tucked 
away under the anterior lobe of the liver, and in getting at it some vessels are certain 
to be cut, and the injection bleeds. I therefore determined to try an injection by the 
portal vein, which is very easy to get at and to ligature, and to expose which only a 
very slight operation is necessary. To my great surprise, the result was a complete 
injection of the veins of the whole body, with apparently nothing whatever in any of 
the arteries. The liver and gall bladder are injected first (and presumably also the 
anterior portal vein). Then the sinus venosus fills up from the liver, and finally the 
injection travels back along all the veins, which become well filled. Examination of 
the specimen afterwards with the dissecting microscope revealed a little injection in 
the auricle, still less in the ventricle, a trace in the cardiac aorta, and none in the 
afferent branchial arteries or in the gills. Nevertheless the pleural sacs were full of 
injection, and it had also passed freely into the large sub-dermal sinus. In this case 
there was and could be no question of rupture, since the injection had first to pass 
through the liver before reaching the sinus venosus. Therefore the fact that the 
pleural sacs filled up in a case where the veins were also well injected, and the possi- 
bility of injection reaching the sacs via the arteries is absolutely out of the question, 
shows that the sacs must at some place be in communication with the venous as well as 
the arterial system. I have not specially investigated this connection, but I believe it 
to be with the superior jugular veins in the neighbourhood of the heart. 

A brief note by Klinckowstrom * may be quoted in confirmation of the above : — 
" Auch im Innern des Korpers bestehen weit ausgedehnte Lymphraume, die sich iiber 
und unter dem (Esophagus, zwischen den ausseren und inneren Zungenmuskeln iiber 
dem Gaumen und rings um das Nasenrohr erstrecken. Wie oben erwahnt, stehen sie 
in weitoffener Verbindung mit den subcutanen Lymphsacken. Audi die Kiemensacke 
sind von zahllosen grosseren und kleineren Lymphraumen, die alle mit einander in 

* lliol. Foren. Forhand., Bel. iv., 1891. 



OiST THE GENERAL MORPHOLOGY OF THE MYXLNOID FISHES. 225 

Verbindung stehen, durchsetzt. Zwischen diesen Theilen des Lymphsystems und dem 
' Pfortaderherzen ' scheint eine Verbindung zu bestehen ; aber nur zwei Male war es 
K. gelungen diese Verbindung durch Injectionen darzulegen ; bei dem einen Exemplare 
hatte die lnjectionsmasse von den Lymphraumen aus nicht nur das Pfortaderherz 
selbst sondern aucli einen Theil des Venensystems gefiillt." 

Klinckowstrom's injections are not as satisfactory or as conclusive as the one 
described above, since he injected not from the portal vein but from the lymph space. 
The fact that the injection reached the portal heart agrees with my results, because the 
mass would reach the latter via the anterior portal vein ( = the right superior jugular). 

3. The Structure of the Vascular Papilla. 

The vascular papillae, although never entirely absent, vary very greatly both in 
occurrence and structure, and they may or may not open into the surrounding sinus. 
It seems obvious, therefore, that they must be regarded as vestigial structures, and that 
they stand for the vanishing point of the connection between the arteries and the 
lymphatics. 

In figs. 2 to 5 I have represented a series of simple cases in which there was an 
undoubted communication between the cavity of the artery and the surrounding lymph 
sinus. They are all drawn from transverse sections of the arteries, but in fig. 5, owing 
to the greater magnification, only the papilla itself and a small portion of the arterial 
wall are shown. Figs. 2, 3 and 5 are sections of afferent branchial arteries, but fig. 4 
is from the common carotid. The spaces occupied by red blood are indicated by the 
red tint. 

In all cases the base of the papilla is excavated as a large space in free communi- 
cation with the cavity of the artery and which therefore always contains blood. If this 
space communicates with the surrounding lymph sinus at all it may do so in a variety 
of ways. The apex of the papilla may be more or less pointed and possess a single 
opening only, as in fig. 3. This is unusual. On the other hand, the free end may be 
expanded, and its external surface indented by a series of crypts, the whole having a 
somewhat digitate or even grape-like appearance, with perhaps one opening or more on 
the summit of each finger. Care must be taken not to confuse these crypts with the 
blood vascular tubules which pass from the basal excavation to the apex of the papilla. 
The latter tubules are usually just large enough to transmit a single corpuscle at a time 
in preserved material, but in the living animal they can be considerably distended. 
Each papilla may contain only one such tubule, as in fig. 3, or several, as in fig. 5, 
and their relation to each other and manner of opening, when they do open, are very 
variable (cp. fig. 4). The papilla may hence be a simple structure with no more than 
a single opening, or a somewhat complex feature with several openings. 

The afferent branchial arteries on rare occasions give off small vessels to the sur- 
rounding tissues before entering the gills, and in one injection a conspicuous twig arose 



226 PROFESSOR FRANK J. COLE 

from the second afferent branchial and broke up on the wall of the corresponding 
efferent gill duct. 

I have never found the vascular papillae entirely absent, but they may be very 
greatly reduced and almost all of them imperforate, in which case the peribranchial sacs 
contain less blood than usual. On the other hand, in those cases where the papillae are 
numerous and well developed, the pleural sacs are always rich in blood. 

Apart from the fact that sections of injected material demonstrate in the clearest 
manner that the vascular papillae may place the cavity of the artery in communication 
with the surrounding "lymph" space, we find in many cases which have not been 
injected a papilla capped by a small blood clot. This clot in favourable cases will be 
found to be directly continuous with the blood coagulum in the interior of the papilla, 
and this in its turn with the contents of the arterial trunk itself. On the other hand, 
in many cases it is equally certain that the papillae are quite closed, and represent 
merely small evaginations of the arterial wall. Occasionally one finds blood corpuscles 
in the tubules, in their external openings, and just outside those openings. 

In injections the mass always passes readily into the basal cavity of the papilla, and 
often into the proximal ends of the tubules ; but it does not always travel along the whole 
length of the tubule, and so to the exterior, or, if it does, the tubule, being elastic, con- 
tracts and empties itself when the pressure is relaxed. When such a papilla is examined 
as a whole preparation, the uninjected distal portion of the tubule is visible as the 
direct continuation of the injected proximal portion. 

A striking feature is that the tubules leading from the large space at the base of the 
vascular papilla to its apex, and it may be opening there, are lined by a well-marked 
epithelium. This epithelium is confined to the tubule, and its extent is indicated by 
the dots in figs. 2 to 5. Peripherally, it does not extend beyond the external opening 
of the tubule, nor does it line the large blood space at the base of the papilla. The cells 
are shallow, but the nuclei are large and deeply staining, and reach from one extremity 
of the cell to the other. 

The following variations in the structure of the papillae may be recorded : — 

1. The tubules traversing the distal end of a papilla may' be so complex, as shown 
by injections, as to almost deserve to be referred to as a rete mirabile. 

2. A papilla may consist of a pear-shaped mass, with its narrow proximal end ex- 
cavated into the usual blood space in free communication with the cavity of the artery. 
At the expanded distal end is a large cavity with very little obvious contents, and 
having no communication with the exterior, but which is connected with the proximal 
blood space by one or more tubules. In other cases the distal cavity was lined by 
epithelium and possessed undoubted openings on to the exterior. 

3. In some of the gelatine injections the papilla was pear-shaped, its enlarged distal 
extremity having a spongy consistency. From the proximal blood space a number of 
fine anastomosing tubules were given off, traversing the spongy mass, and some of them 
opening by fine pores on to the exterior. 



ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 227 

4. A papilla may be completely injected, so that the capillaries within the contiguous 
walls of the artery are filled, without, however, any signs of the injection escaping from 
the papilla. In these cases the interior of the papilla is completely occupied by the 
blood space, and there is no distal spongy end. 

5. In one injection the free spongy extremity of the |)apilla was traversed by several 
of the blood vascular tubules, some of which were observed to unite to form a large 
vessel opening on to the exterior by a small pore. 

6. In another injection the stalk of the papilla contained the customary large 
simple blood cavity, from which there were given off distally a number of tubules, the 
latter combining in the distal bulbous end of the papilla to form an elaborate vascular 
network. 

7. From the large central blood space of the papilla there may pass into the more or 
less solid tissue of the free end a number of intercellular blood canaliculi. These differ 
markedly from the much larger and well-defined blood vascular tubules, which may also 
be present. They cannot be described as blood capillaries, since they have no definite 
walls, but rather resemble the bile capillary. 

4. Theoretical Considerations. 

The development of the lymphatic system has of late years been successfully 
studied by several American morphologists. The most recent paper on the subject 
is by Huntington on the development of the lymphatic system in Reptiles.* Ac- 
cording to this author, the systemic lymphatic channels arise by the confluence 
of mesenchymal spaces independently of the blood vascular system. They are 
not derived from veins, their cavities are independent of those of the veins, and they 
are lined with a " lymphatic vascular endothelium not derived from or connected with 
a pre-existing haemal vascular endothelium." 

On the other hand, the jugular lymph sac, both of Reptiles and Mammals, has an 
altogether different origin. This arises by the fusion of a venous plexus, directly 
derived from the venous channels of the pericardial area. The endothelium of the veins, 
therefore, is continuous with that of the lymph sinus, and the latter at first contains 
red blood. " It then evacuates its early blood contents, separates temporarily from the 
haemal vascular system, establishes secondary connections with the systemic lymphatic 
vessels of the anterior part of the body and of the anterior extremity, and finally re- 
enters the venous system at a definite and constant point of secondary lymphatico- 
venous junction." 

We are unfortunately ignorant of the development of the peribranchial sacs and other 
so-called lymph spaces in the Myxinoid fishes, but the facts disclosed by the American 
morphologists suggest a possible explanation, or at all events indicate a line of inquiry. 
If the jugular lymph sac of Reptiles and Mammals is formed by the fusing up of a 
venous network, and a sinus thereby formed, it may be that the peribranchial sacs of 

* Anat. Rec, June, 1911. Cp. also January, 1910, and Mem. Wistar Inst. Anat. Biol., No. I., 1911. 



228 PROFESSOR FRANK J. COLE 

the Myxinoid are also formed by the coalescence of a vascular network. On this view, 
the vascular papillae will represent the remains of the embryonic plexus. In the closed 
papillae the detachment of their contributions to the sinus will have been complete, 
whilst the open papillae will have retained the embryonic continuity of the two systems 
of spaces. 

As regards Petromyzon there are some grounds for this view. Cori points out 
that the branchial sinuses are represented in earlier stages by what he calls a venous 
network. Schneider also has the following significant passage : " Zwischen der 
Epithelialschicht der Mundhohle und der Muskulatur de Korperwand befindet sich im 
Ammocotes eine Schicht sogennanten adenoiden (cytoiden Binde-) Gewebes, welches 
von zahlreichen Capillaren durchsetzt wird. Beim Uebergang in den Petromyzon 
nehmen diese Capillaren an Grosse zu und verschmelzen zu einem grossen Venensinus." 

The fact that all the so-called lymph spaces in Myxinoids normally contain blood, 
though in varying quantities, must be held to remove these spaces from the category 
of lymph spaces sensu stricto. At the same time, however, they cannot be said to lie 
in the direct course of the blood stream, and for this reason must be equally excluded 
from the blood vascular system. Langerhans declines to commit himself as to what 
system the spaces belong to, but later writers have generally referred them to the 
lymphatic system, and denied that blood is normally present in them. Mozejko, 
however, takes the intermediate course, and regards the spaces as blood-vessels and 
lymphatics at the same time. 

There is, to my mind, little doubt that as regards the vascular system the Marsipo- 
branch fishes have reached the parting of the ways. The blood system is well developed, 
but not completely developed. There is still in places an ill-defined connection between 
the arteries and the veins. Similarly, the lymphatic system is indicated as far as its 
broad outlines are concerned, but it contains red blood, and has not yet acquired its 
independence. It is, in fact, in the act of becoming detached, as is illustrated by the 
great variation in the amount of blood it contains, and in the fluctuating extent to 
which it receives blood from the arteries via the vestigial vascular papillae. 

Summary. 

1. Red blood occurs normally, but to a variable extent, in all the so-called 
" lymphatic " spaces of Myxine. In the living animal such blood can readily be seen 
in the extensive sub-dermal sinus. 

2. In the region of the gills, and especially in the case of the peribranchial or pleural 
sacs, red blood enters the spaces from the arteries. This has been actually observed 
under the microscope in certain injection experiments on the freshly killed animal. 
The means of communication between the arteries and the sacs are a number of 
perforated papillae situated on the afferent and efferent branchial vessels. The blood 
so entering the sacs is drawn off again into the venous system via the jugular veins. 



ON THE GENERAL MORPHOLOGY OF THE MYXINOID FISHES. 229 

3. Only a variable proportion of the vascular papillae are perforated. Many, if not 
most, of them are closed and vestigial. 

4. The "lymphatic" spaces are neither true lymphatic nor true blood vascular 
spaces, but j>artake of the nature of both. We have in the Myxinoids the final stage 
in the separation of the blood vascular from the lymphatic system. In other words, 
the two systems are in the act of segregating out. 

I am indebted for the expenses of this research to a grant from the Government 
Grant Committee. 

It gives me pleasure to acknowledge the kindness of Professor A. Meek, M.Sc, in 
placing the resources of his admirable laboratory at Cullercoats at my disposal, and to 
the assistant naturalist, Mr B. Storrow, for much valuable assistance in carrying out 
the injection experiments. 

October 1911. 

Postscript, January 1912. — By the courtesy of Dr B. Mozejko of Warsaw, I have 
received advance proof sheets of a paper, which by this time will have been published, 
on the morphology of the vascular system of the Lamprey. Tn this paper Dr 
Mozejko states that the contents of the gill sacs cannot be distinguished from venous 
blood, and he believes the sinus system to be formed during metamorphosis by the 
local enlargement of vascular networks. The sinuses are not present in the young 
Ammoccetes, and only appear shortly before metamorphosis. He concludes that the 
venous system of Petromyzon is not a venous system sensu stricto, but is a 
" sy sterna venoso-lymphaticum," in which the venous and lymphatic channels are fused 
into a common system, and in which the lymphatic vessels have only partially 
segregated out. These conclusions are very similar to those I had already arrived at 
from a study of Myxine. 



EXPLANATION OF THE PLATE. 

Fig. 1. Reconstruction, from serial sections, of the gills, branchial oesophagus, and afferent and efferent 
branchial arteries (coloured red) of a 25 cm. Hag, seen from the left side, x 11. The muscles of the gut 
are not included. Afferent and efferent gill ducts are represented cut off at their origin. In the case of the 
common carotid, the branches of the artery are distinguished from the vascular papilla? by a cut extremity. 
Some of the papillae on the afferent branchial arteries are compound, but the details cannot be shown in a 
diagram of this magnification. For the same reason two very small papillae on the first afferent branchial, 
one on the second, and two on the fourth, are omitted. 

Fig. 2. Transverse section of an uninjected afferent branchial artery bearing a vascular papilla with two 
openings into the pleural sac. x 50. The division between the artery and the papilla is indicated by the 
broken line. The extent of the epithelial lining of the tubules leading to the exterior is shown by dots. 
Blood vascular cavities coloured red. The drawing is diagrammatic to the extent that both openings did 
not appear in the same section, 



230 



ON THE GENERAL MORPHOLOGY OF THE MYXINOLD FISHES. 



Fig. 3. Transverse section of an uninfected afferent branchial artery with a vascular papillae having only 
one opening (very unusual). x 50. The tubule leading to the surface and the opening itself are distinctly 
larger than is customary. For description see fig. 2. 

Fig. 4. Transverse section of the common carotid artery in the region of the gills showing two vascular 
papillae. x 50. The figure is slightly diagrammatic in so far as though both vascular papillae appear in the 
same sections, the openings of the lower papilla, as figured, were in the next section but one to that drawn. 
The upper papilla had two openings, and two tubules which ended blindly. The lower papilla had two 
openings. The specimen had been injected with soluble Prussian blue. Cp. the description of fig. 2. 

Fig. 5. Portion of an afferent branchial artery with one vascular papilla mounted entire and examined 
as a transparent object, x 100. Injected with carmine gelatine. Only a very small portion of the wall of 
the artery is shown. Cp. the description of fig. 2. 

Reference Letters. 



af. br. Afferent branchial artery. 
a. g. d. Afferent gill duct. 

a. s. ao. Anterior or pre-branchial portion of the 
dorsal or systemic aorta. 
au. Auricle of the heart. 
br. cl. Branchial cloaca. 
c. ao. Ventral or cardiac aorta. 
c. car. Common carotid artery. 



d. oes. ct. Ductus cesophago-cutaneus. 
ef. br. Efferent branchial artery. 
e. g. d. Efferent gill duct. 

oes. Oesophagus — branchial portion. 
s. ao. Dorsal or systemic aorta. 
s. v. Sinus venosus. 
vent. Ventricle of the heart. 



PR] 



, 



an mo; 





Roy. Soc. Edin r - 

Prof. F. J. Cole on the Morphology of Myxine. Part IV. 



Vol. XL VIII. 




Fig. 2. x 50 




Fig. 3. x 50 




ig. 4. x 50 




'g- 5- x 100 







M'Farlane & Erskine Edir 



■ 

% ■ 



( 231 ) 



XII. — The Effect of changing the Daily Routine on the Diurnal Rhythm in Body 
Temperature. By Sutherland Simpson, M.D., D.Sc. (From the Physiological 
Laboratory, Medical College, Cornell University, Ithaca, N.Y., U.S.A.) (With 
Thirteen Figures in the Text.) 

(MS. received November 20, 1911. Read January 8, 1912. Issued separately April 4, 1912.) 

CONTENTS. 

PAGE 

I. Introduction ............. 231 

II. Previous Work on the Subject .......... 232 

III. Present Investigation ........... 236 

IV. Results and Discussion ........... 246 

V. Effect of Muscular Rest and Activity on Body Temperature ..... 250 

VI The Relative Values of Records from Rectum, Mouth, and Axilla, as indicating Changes 

in Body Temperature ........... 257 

VII. Summary ............. 260 

I. Introduction. 

It is a well-established fact that the temperature of the human body is not con- 
stant, but shows a distinct and fairly regular daily rhythm. In the classical curves of 
Jurgensen and Liebermeister * the minimum is reached in the early morning, some time 
between the hours of 2 and 6, and the maximum in the late afternoon or evening, 
between 4 and 8. In both there is a sharp morning rise, followed by a slight fall in 
the late forenoon or early afternoon, and again a further rise to the maximum, which 
is reached some time between the hours stated. It has been demonstrated repeatedly 
by many subsequent observers that this general type of curve is a more or less correct 
representation of the changes which the rectal temperature undergoes, in the course 
of the twenty-four hours, in a healthy individual leading an active life during the day 
and sleeping at night. 

While the presence of this daily rhythm is well recognised, the causes underlying 
it are not clearly understood. The factors which are believed to affect the body 
temperature in health are muscular exercise, mental effort, ingestion of food, light, 
temperature of the surrounding medium, and sleep. Of these the first is the most 
important. Each and all of these influences are active in the body at different periods 
throughout the twenty-four hours, but the exact relationship existing between them 
and the diurnal rhythm in the body temperature has never yet been satisfactorily 
demonstrated. 

By many it is held that the combined action of the several influences enumerated 

* Jurgensen and Liebermeister, Handbuch der Pathologie und Therapie des Fiebers, Leipzig, 1875. 
TRANS. ROY. SOC. EDIN, VOL. XLVIII. PART IT. (NO. 12). 36 



232 DB. SUTHERLAND SIMPSON ON 

above, although it may modify it to a considerable extent, cannot account entirely for 
this diurnal fluctuation. They believe that there exists in the body a fixed periodicity 
of which the temperature rhythm is an expression, and that this periodicity persists 
under all conditions, and is, to a large extent, independent of outside influences. Others 
are inclined to question the existence of this mysterious periodicity, and to look upon 
the diurnal variation as being due entirely to the action on the body of the influences 
already mentioned, which are known to raise or lower the temperature. 

IT. Previous Work on the Subject. 

If these factors are alone responsible for the diurnal temperature wave, then any 
modification in the application of them, and particularly any change in the time 
within the twenty- four hour period at which one or more of them operate, might be 
expected to produce a corresponding change in the temperature curve, or, in the 
simplest and most complete case, total inversion of the daily routine should cause a 
complete inversion of the temperature curve. 

Various attempts have been made by different investigators to produce an inverted 
day-and-night curve, but these have been unsuccessful in the human subject. It is 
true that the earlier observers, Debczynski, # JAEGER,t and Buchser,| claim to have 
succeeded, but when the conditions under which their experiments were carried out 
come to be examined, it is found that the methods of all three are open to criticism, 
and that their results are inconclusive. 

More recently Mosso § (1885) and Benedict || (1903), in carefully planned experiments, 
have studied the effect of night work and day rest and sleep on the temperature rhythm, 
and both have arrived at the conclusion that the normal temperature curve cannot be 
inverted by inverting the daily routine, for while rest and sleep during the day lower 
the temperature, work during the night does not appreciably raise it. The curve is 
modified by the altered conditions but it is not reversed. 

These results then would appear to show that the daily oscillations of the body 
temperature are not due directly to the causes already mentioned, but that they may 
have a deeper significance, and may indicate a diurnal periodicity in the body 
comparable in character to the seasonal and lunar changes that are known to occur 
in certain plants and animals. That the temperature rhythm is to some extent fixed 
in the body is the most obvious interpretation to be put upon the failure of Mosso and 
Benedict to invert it, and this apparent fixity of rhythm is difficult to explain. As 
Benedict says : " Why the temperature of the human body reaches a minimum at 
2 a.m. to 6 a.m., independent of whether the subject is sleeping soundly in the 

* DEBCZTN6KI, Abstract in Jahresb. der ges. Med., Bd. x., 1875, p. 248. 

t Jakgbr, Deutsches Arch. f. lclin. Med., Leipzig, Bd. xxix., 1881, p. 533. 

j Buchser, quoted by Carter, Jour. Nerv. and Merit. IHs., vol. xvii., 1890, p. 785. 

§ Mosso, Archives italiennes de biol., vol. viii., 1887, p. 177. 

|| Benedict, Amer. Jour, of Physiol, vol. xi., 1904, p. 143. 



Daily routine and body temperature. 233 

recumbent position or whether he is awake and sitting, or even standing and walking, 
is a problem that calls for extended research." 

Instead of altering the daily routine artificially in a fixed locality, the same result 
may be effected in a natural way by changing the locality, for an individual who 
travels round the world in these days of rapid transit from West to East, or vice versa, 
especially in high latitudes, quickly changes his daily routine. The happy idea of 
applying this method to the study of the question under discussion seems first to have 
occurred to Gibson.* The opportunity presented itself when he had occasion to make 
a voyage from New Haven, Connecticut, across the American continent and Pacific 
Ocean to Manila in the Philippines. As a difference of eleven hours exists between the 
local times of these two stations, the journey involved the shifting of the daily routine, 
so that day and night were practically reversed. 

Gibson's observations were made on himself and a second subject at different 
stages of the voyage, and on individuals in Manila who came originally from the 
United States and who had resided in Manila for varying periods of time. He made 
control experiments on his own body temperature for two days before leaving New 
Haven, taking readings every two hours, and so obtained the diurnal temperature 
curve, which proved to be of the ordinary type commonly described as normal. The 
same process was repeated several times during the voyage and again at Manila. 

He found that the " transposition of the daily routine through a period of nearly 
half a day, experienced as the result of the time changes during the trip from New 
Haven to Manila, was accompanied by an immediate adjustment of the rhythmic 
temperature variation to the new regime in the case of the writer and of a second 
subject, so that on arrival in the islands the curves obtained were still normal in 
character. Subsequent residence in the Philippines for a period of about six weeks 
induced no alterations of any significance. Observations made during the return trip 
showed an apparent adjustment of the temperature rhythm day by day coincident with 
the shifting of the routine. After returning to New Haven the record continued to be 
normal, and closely resembled the earlier controls. . . . Additional observations on 
individuals who came originally from the United States and who have resided in Manila 
for varying periods of time, corroborate in part the results on the writer, in so far as 
the temperature rhythm was found to be of the ordinary type." 

A similar experiment was made by Osborne t three years later, apparently without 
any knowledge of the previous work of Gibson. Having first noted that his own daily 
maximum temperature in Melbourne occurred about 6 p.m., while on a voyage from 
that city to London he made some observations on his own rectal temperature with the 
object of seeing whether this maximum remained throughout the voyage at 6 p.m. 
according to Melbourne time or took place at a certain hour relative to local or ship's 
time. The difference in longitude between the two places is equivalent to about ten 

* R. B. Gibson, Amer. Jour, of the Med. Sciences, June 1905, p. 1048. 
t Osborne, Journal of Physiology, Proceedings, Jan. 25, 1908. 



234 DR SUTHERLAND SIMPSON ON 

hours in time, and if the Melbourne rhythm had persisted, in London the curve would 
be practically reversed. Osborne found, however, that his maximum tended to follow 
local time, occurring about 6 p.m. by the ship's clock. 

But he was only able to make five sets of observations, and in one of these the 
highest temperature was reached at 10.15 a.m. by ship's time, corresponding to 5.52 p.m. 
Melbourne time. In another only four readings were taken, afc hours which did not 
cover the 6 p.m. by Melbourne time (2 p.m., 5.30 p.m., 6 p.m., and 11 p.m. ship's 
time corresponded to 0.48 a.m., 4.18 a.m., 4.48 a.m., and 9.48 a.m. Melbourne time), 
so that it is impossible to say whether the maximum agreed with the latter or not. 
The results therefore are not quite convincing. To quote his own words : " The above 
results, though incomplete, rather tend to prove that the time of evening maximum takes 
place with regard to local time and not the time of the starting-point — Melbourne. 
They do not, however, disprove the existence of body periodicity, nor prove that the 
evening maximum is determined solely by the hours of sleep, the activities of the day, 
and the diurnal variations of light and heat, for a true periodicity might have been 
present but adjusted to the new conditions owing to the very gradual manner these 
were introduced." * 

The latest contribution to the subject comes from LiNDHARD,t the medical officer of 
the Danish Arctic expedition to the north-east coast of Greenland in the years 1906-8. 
On this expedition he studied extensively, on himself and other members of the ship's 
company, the effects of various conditions on the rectal, mouth, and skin temperatures. 
Discussing the question of periodicity in relation to body temperature, he criticises the 
method of Mosso and Benedict, as Tigerstedt and the present writer had done before, 
on the ground that when one attempts to reverse the daily routine in a single 
individual by arranging that he shall work during the night and rest and sleep in the 
day-time, one cannot disconnect this individual from the rest of society. So long as 
the society, of which the individual is a part, follows a fixed rotation, the latter, 
consciously or unconsciously, will tend towards the same mode of life. Under reversed 
day-and-night conditions the subject must necessarily work with artificial light in the 
stillness of the night, and he must sleep in the day-time through noise and other dis- 
turbing influences. The night-worker is in sympathy with a sleeping world, and his 
bodily activities are unconsciously affected by this circumstance. 

Relating his own experience in night-watching, he calls attention to the fact that 
the "night time has a peculiar effect upon one's general state of mind owing to the 
stillness, solitariness, and various other circumstances, and this changed psychical 
condition reacts on all one undertakes. Every strong or sudden noise is disagreeable, 

* In a recent paper (Report of the Danish Expedition to the North-East Coast of Greenland, 1906-8, vol. xliv., 
1910, p. 1 ; Reitzel, Copenhagen) Lindhard, unaware of Gibson's work, gives to Osborne the credit of being the first 
to adopt this method of changing the daily routine, and makes no mention at all of Gibson's name in this connection. 
As a matter of fact priority, by three years, belongs to Gibson, and besides, his observations were much more 
extensive and complete, and his results far more definite than those of Osborne. 

t Lindhard, loc. cit. 



DAILY ROUTINE AND BODY TEMPERATURE. 235 

and involuntarily one tries to avoid all such things ; the movements become gliding and 
noiseless, resembling those of other night animals, and are much less definite than in 
the day-time. A round in the night will therefore fail to raise the temperature like a 
corresponding walk in the day-time. And one is never sitting so still as in the 
night; . . . ." 

The psychical factor might naturally be expected to have a greater influence on the 
human subject than on the lower animals, and this may account for the fact that in the 
monkey the body-temperature curve can be reversed by reversing the daily routine 
(Simpson and Galbraith*). Here, however, the case is somewhat different in another 
respect, since the reversed conditions were imposed on a colony of monkeys instead of 
an isolated individual. 

Instead of experimenting on a single individual, then, the proper method would be 
to apply the "reversed" routine to the whole society in which the individual lives, but 
this is inconsistent with social life in civilised communities. Such an experiment, how- 
ever, may be carried out on a high-arctic expedition during the long polar night or day, 
and this Lindhard did on the members of the party. For this he preferred the winter 
night to the perpetual day of summer, since the " sun occasions a continual unrest, the 
working hours are unlimited, and a little food emancipates one from the regular meals. 
It will therefore be very difficult for all to adopt the same mode of life, .... Apart 
from this, the difference in the intensity of the light is much more pronounced in 
summer than in winter, just as the variations in the day-and-night temperature are 
much greater. During the polar night the case is different. All lamps are put out at 
certain fixed hours appointed by the leader of the expedition ; then it is night : during 
the remaining hours the lamps are lit and we have day. As practically all work is done 
indoors, the dining hours are kept by everyone ; and in the open air the light by day 
and at night does not vary appreciably, and the changes in temperature are com- 
paratively small." 

This experiment was carried out in January 1907, in lat. 76° 46' N., on the whole 
ship's crew. The transition was made very quickly and easily. Bedtime was delayed 
once four hours, and then eight hours by the insertion of an extra meal, and so in two 
days the "reversal" was accomplished. The lamp was lit and extinguished at a fixed 
time as before, but twelve hours later, and all met at the table as usual at the 
" reversed " meal hours. Only on clear days at " noon " a little brightness in the 
southern sky might be missed, otherwise the change was betrayed by no external sign. 
All were conscious of the fact that this change had been made, but in the great 
majority this did not give rise to a feeling of anything unusual. " Time rolled along 
on its even, ordinary way. More than half of the twenty-eight members of the 
expedition felt, indeed, just as usual as soon as the transition had been accomplished ; 
after five to six days only a few were a little indisposed to work, sleeping less well in 
the " night " and becoming sleepy at various times of the " day." The function altered 

* Simpson and Galbraith, Tram. Roy. Soc. Edin., vol. xlv., part i., 1905, p. 65. 



236 DR SUTHERLAND SIMPSON ON 

with the greatest difficulty was defalcation ; for a few it took about a week before this 
occurred at the usual time, for one (rarely quite free, however, from indisposition) 
it took even till the end of the experiment." The return to the normal routine 
was made in the same, way as the reversal. The period of complete reversal lasted 
eleven days. 

Although readings for the seven hours of sleep were not taken, the curves of all 
point to reversion. In those who had difficulty in becoming accustomed to the altered 
routine there is delay in the adaptation of the temperature curve to the changed 
conditions, but ultimately " in all cases the fundamental type is evident, and the causes 
of the departures present are obvious. All of them tend to show that the curve of 
temperature variations is determined by work and mode of living, that the astronomical 
division of day and night is without importance in this regard, and that an inherited 
form is consequently out of the question, a mysterious periodicity even more so." 

III. Present Investigation. 

The writer happened to be present when Professor Osborne made his communication 
at a meeting of the Physiological Society in London on January 25, 1908, and being 
especially interested in the subject of body temperature, he was struck by the fact that 
no one had ever thought of adopting the procedure described by Osborne before. At 
that time it did not seem probable that the opportunity would ever present itself to 
the writer of repeating the work of Osborne, which was admittedly somewhat frag- 
mentary and inconclusive, but circumstances came about which made this possible a 
year later. 

In the summer of 1909 a journey was undertaken from Ithaca, in the western part 
of the State of New York, eastward to Edinburgh, and after a six weeks' stay in 
Scotland, westward again from Edinburgh to Winnipeg. As the local time of Edinburgh 
is about five hours in advance of Ithacan time and over six hours ahead of Winnipeg 
time, in a rapid journey between these places the daily routine would, within a few 
days, undergo a considerable modification although it would not be completely reversed. 
At that time I did not know of Gibson's work, nor, in fact, until I arrived in Edinburgh, 
when I had the privilege of reading his paper. 

It was planned to make, as far as possible, three-hourly observations on the 
temperature of the rectum, mouth, and axilla during the waking hours throughout the 
voyage. In order to obtain the daily temperature curve in Ithaca for comparison, the 
observations were begun one week before starting on the journey eastward. It was 
found that the steepest grade on the curve took place between 6 a.m. and 9 a.m., and 
in this interval readings were taken more frequently. The subject (the writer 
himself) was forty-six years of age, measured 5 feet 10|- inches in height, weighed 
205 pounds, and was of stout build. He was in perfect health throughout the 
whole experiment. 



DAILY ROUTINE AND BODY TEMPERATURE. 237 

For the mouth and axilla sensitive and accurate clinical thermometers were used 
(Fahrenheit scale), graduated in fifths and capable of being read to tenths of a degree. 
For the rectum a special thermometer was constructed, also of the clinical type, but 
the bulb and about 1 cm. of the stem was encased in silver as a safeguard against 
accidental breakage while in use on shipboard. In this the centigrade scale was 
adopted ; each degree was subdivided into tenths, but with a lens readings to the 
second decimal place could be made with approximate correctness. 

The records were taken simultaneously from the left alveolo-lingual sulcus, the 
Left axilla, and the rectum, where the thermometer was always inserted to the same 
depth — 8 cm., — as indicated by a fine wire fixed around the stem at this point. 
The thermometers, before being used, were warmed in the hand, and they were 
held in position for five minutes, although it is probable that a shorter time would 
have been sufficient for the mercury to adjust itself to the temperature of the 
surrounding walls. 

For comparative results of this kind it is essential that the daily habits of the 
subject should be, as far as possible, the same throughout the whole experiment, and 
this was kept in mind. Three meals a day were taken at the same hours — breakfast 
between 8 and 9 a.m., lunch from 1 to 2 p.m., and dinner between 6 and 7 p.m. No 
food was taken from 7 p.m. till 8 a.m. except very occasionally, when a glass of 
buttermilk with a slice of bread was served for supper about 9 o'clock. 

During the control period in Ithaca the days and evenings were spent, for the most 
part, in the laboratory, — sometimes seated at a table writing or reading, sometimes 
walking or standing about in the rooms doing light work. The subject got out of bed 
about 7 a.m., dressed, and walked about three hundred yards to breakfast, and then up 
a hill, which was fairly steep, about half a mile to the laboratory. The temperature 
was taken at 6 a.m., frequently again before arising about 7, after dressing about 8, and 
always at 9 a.m. in the laboratory. On warm days, instead of walking, the street car 
was used. Lunch was taken in the laboratory and dinner at the same dining-rooms as 
breakfast, after which the street car was made use of again in returning to the laboratory. 
Some time between 10 and 11 p.m. a walk of ten minutes down the hill brought the 
subject to his living-rooms, and he retired about 12, after the last set of observations 
for the day had been recorded. The daily bath was taken at night instead of in the 
morning during this period. Occasionally both in the control and other periods, when 
the subject chanced to wake up in the early morning, a set of readings was taken, but 
this practice, of course, could not be carried out regularly. 

The daily routine above described is not very different from what can be followed 
on board ship, and as far as possible both there and in Scotland Ithacan habits were 
kept up. For example, before breakfast (the meals were served at the same hours as 
in Ithaca) a short walk was taken around the deck in imitation of the walk to the 
dining-rooms in Ithaca, and again after breakfast. The day was spent for the most 
part in reading novels in the saloon or on deck, promenading, and playing deck billiards 



238 DR SUTHERLAND SIMPSON ON 

or quoits. The muscular exercise indulged in was probably not equal to, and certainly 
not in excess of, that taken in Ithaca, but there was greater mental relaxation. The 
weather was fine from land to land and not the slightest suspicion of sea-sickness was 
experienced. 

The eastward voyage was begun on June 25, about 11.45 p.m., when the train left 
Ithaca, and New York City was reached shortly after 8 o'clock next morning. The 
steamer (s.s. Caledonia, Anchor Line) sailed from New York for Glasgow about 
2.30 p.m. on June 26. Four observations were made on the train and two in New 
York City for the sake of keeping up the continuity. Glasgow was reached about 

4 p.m. and Edinburgh at 9 p.m. on July 4, the voyage thus lasting practically 
eight days. 

The longitude of Ithaca, N.Y., is 76° 29' W., and the local time is therefore 5 hours 

5 minutes and 56 seconds — practically 5 hours 6 minutes — behind Greenwich time. 
The clock time at Ithaca is about 6 minutes ahead of the true local time. Eastern 
standard time is used at Ithaca, and this is the local time for stations on long. 75° W., 
which, is 5 hours behind Greenwich time. My watch was adjusted for this difference, 
and the observations were made according to the correct local time at Ithaca. The 
longitude of Edinburgh is 3° 10' W., and the local time about 12 minutes behind 
Greenwich time, so that the difference between Ithaca and Edinburgh local time is 
4 hours 54 minutes. 

On the Atlantic voyage the ship's position at noon was posted each day, and from 
the longitude given the local time was obtained. This was only correct for noon, how- 
ever, but it was known that the time gained in a day's run east would be a little over 
half an hour, and, assuming that the vessel's speed was uniform, allowance could be 
made for the gain in the interval between each set of observations. This only amounted 
to a few minutes, however, and for my purpose it was immaterial whether this error 
was accurately corrected or not. The readings were taken usually at the hours stated, 
but on some occasions this was not possible and a variation of ten minutes to one side 
or the other was not considered sufficient to affect the results appreciably. 

The figures obtained for the rectal, mouth, and axillary temperatures are given in 
degrees centigrade in the subjoined table, together with the temperature of the air, the 
local time, and, after the eastward journey was begun, the corresponding time at Ithaca, 
the starting-point. The remarks, necessarily brief, may be sufficient to show how the 
subject was employed for some time immediately preceding the observations. 



DAILY ROUTINE AND BODY TEMPERATURE. 



239 



TABLE I. 



Records of Rectal, Mouth, and Axillary Temperatures taken at Ithaca, N.Y., on the Journey from 
Ithaca to Edinburgh, and for Six Days after arriving in Edinburgh. The Air Temperature, 
Ithaca Time and Local Time are also given. 





Ithaca 
Time. 


Local 
Time. 


Rectum. 


Mouth. 


Axilla. 


Air. 


Remarks. 


1909 
















June 19 


6 a.m. 




371 


363 


362 


23° C. 


Awake since 5.30 a.m. Temp, taken in bed. 




9 „ 




374 


36-8 


361 


24 


In laboratory, half an hour after breakfast. 




12 „ 




37-4 


36 6 


363 


26 


„ writing at table. 




3 p.m. 




376 


37-0 


36-4 


25 


,, moving about. 




6 „ 




379 


36-8 


363 


24 


)) 15 




9 „ 




375 


368 


362 


24 


„ arranging books. 




12 „ 




37 2 


36-5 


36-1 


23 


In rooms ; just before retiring. 


June 20 


6 a.m. 




36-8 


36 3 


363 


21 


In bed ; just awoke. 




7 „ 




369 


36-2 


363 


22 


„ light sleep since 6. 




8 „ 




37-1 


365 


36-2 


22 


After dressing. 




9 „ 




37-4 


36-8 


362 


24 


In laboratory ; took street car. 




12 „ 




37-6 


37-0 


36-2 


25 


„ doing experiment. 




3 p.m. 




376 


36-8 


363 


26 


»> )' 




6 „ 




37-4 


369 


36-5 


25 


,, cleaning apparatus. 




9 „ 




37-5 


36-8 


362 


26 


In rooms ; reading and writing. 




12 „ 




37-3 


368 


36-0 


24 


„ before retiring. 


June 21 


4 a.m. 




36-8 


363 


36 3 


20 


In bed ; just awoke. 




6 „ 




36-9 


363 


363 


20 


„ asleep since 5. 




7 „ 




370 


36-2 


36-1 


21 


,, light sleep since 6. 




8 „ 




372 


36-8 


362 


22 


After dressing. 




9 „ 




377 


373 


37-1 


25 


After walking to laboratory. 




12 „ 




37-4 


36-9 


364 


28 


In laboratory, writing at table. 




3 p.m. 




37-5 


371 


35-8 


29 


>) 5) >> 




6 „ 




37-6 


369 


36-1 


29 


Shopping down town. 




9 „ 




376 


37-1 


36-3 


25 


In laboratory, writing at desk. 




12 „ 




37-4 


367 


363 


22 


In rooms ; did not retire till 1 a.m. 


June 22 


6 a.m. 




36-8 


36-3 


361 


19 


In bed ; awake since 5. 




7 „ 




36-8 


365 


363 


22 


)> 




8 „ 




37-2 


36-8 


36 3 




After dressing. 




9 „ 




37-4 


372 


367 


24 


In laboratory, by street car. 




12 




376 


37-2 


367 


26 


Had been walking on campus. 




3 p.m. 




37 5 


37-1 


36-4 


25 


In laboratory ; cleaning lenses. 




6 „ 




37-0 


36-8 


363 


25 


Writing near open window. 




9 „ 




37-6 


37-1 


366 


25 


In laboratory, after ride on street car. 




12 




37 3 


36-8 


36-0 


23 


In rooms; retired at 12.30. 


June 23 


6 a.m. 




370 


36-5 


36-4 




In bed ; sound asleep since 1 a.m. 




7 „ 




371 


36-6 


364 


25 


,, room hot. 




8 „ 




37-4 


36-9 


36-2 


24 


After dressing. 




9 „ 




37-6 


36-9 


363 


26 


In laboratory, by street car. 




12 




37-6 


370 


36-5 


27 


„ moving about. 




3 p.m. 




37 4 


36-9 


36-4 


27 


>> 5) 




6 „ 




377 


37-1 


36'7 


26 


„ after being down town. 




9 „ 




37-8 


373 


371 


25 


In rooms ; walked home. 




12 




37-4 


369 


36-7 


23 


„ packing trunk. 


June 24 


3 a.m. 




36-7 


36-2 


36-1 




In bed ; asleep since 12.30. 




7 „ 




37-0 


36-6 


36 3 


24 


„ did not sleep well. 




9 „ 




37 9 


37-4 


371 


27 


After walk to laboratory. 




12 




37-9 


37-2 


366 


28 


In laboratory ; in active movement. 




3 p.m. 




38-5 


37-6 


372 


31 


„ after short walk on campus. 




6 „ 




38-1 


37-1 


36-6 


29 


,, moving about actively. 




9 „ 




37-9 


37-0 


36-6 


27 


„ packing books. 




12 




373 


36-6 


36-1 


25 


In rooms ; reading. 


June 25 


2 a.m. 




36-8 


36-4 


363 




In bed ; had been asleep. 




7 „ 




36-7 


36-4 


362 




„ after sound sleep. 




8 „ 




36-9 


366 


36-4 


25 


After dressing. 




9 „ 




37-2 


37-1 


36-6 


25 


In laboratory, by street car. 




12 




38-2 


37-6 


37-2 


28 


Long. 76° 29' W. In lab. ; moving about. 




3 p.m. 




37-8 


37-4 


37-1 


29 


In laboratory ; packing books and papers. 




6 „ 




37-7 


37-4 


37-0 


29 


„ moving about actively. 



TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 12). 



37 



240 



DR SUTHERLAND SIMPSON ON 



TABLE I.— continued. 



1909 
June 25 

June 26 



June 27 



June 28 



June 29 



June 30 



July 1 



July 2 



Ithaca 
Time. 


Local 
Time. 


Rectum. 


9 p.m. 




376 


12 


12' " 


376 


2.56 a.m. 


3 a.m. 


368 


6.51 „ 


7 „ 


37 3 


7.50 „ 


8 „ 


375 


9.50 „ 


10 „ 


376 


11.50 „ 


12 


38-1 


2.50 p.m. 


3 p.m. 


37-8 


5.47 „ 


6 „ 


37-9 


8.44 „ 


9 „ 


37-6 


11.41 „ 


12 


37 3 


5.33 a.m. 


6 a.m. 


36-8 


6.31 „ 


7 „ 


36-8 


7.29 „ 


8 „ 


37-0 


8.27 „ 


9 „ 


37 2 


11.23 „ 


12 


37 3 


2.19 p.m. 


3 p.m. 


369 


5.15 „ 


6 „ 


368 


8.11 „ 


9 „ 


36-7 


11.07 „ 


12 


37-2 


3.01 a.m. 


4 a.m. 


36-4 


5.57 „ 


7 „ 


36 4 


6.56 „ 


8 „ 


368 


7.54 „ 


9 „ 


37 


10.50 „ 


12 


37-2 


1.46 p.m. 


3 p.m. 


369 


4.42 „ 


6 „ 


37-0 


7-37 „ 


9 „ 


37-2 


10.32 „ 


12 


36-8 


5.24 a.m. 


7 a.m. 


366 


6.22 „ 


8 „ 


36-9 


7.20 „ 


9 „ 


37 2 


10.16 „ 


12 


371 


1.12 p.m. 


3 p.m. 


36-9 


4.08 „ 


6 „ 


369 


7 '04 „ 


9 „ 


37-0 


10.00 „ 


12 


36-4 


12.56 a.m. 


3 a.m. 


36-2 


3.52 „ 


6 „ 


36 3 


5.49 „ 


8 „ 


36-5 


6.48 „ 


9 „ 


36-9 


9.44 „ 


12 


36-9 


12.40 p.m. 


3 p.m. 


369 


3.36 „ 


6 „ 


36-8 


6.32 „ 


9 „ 


37-4 


9.28 „ 


12 


373 


3.21 a.m. 


6 a.m. 


369 


4.19 „ 


7 „ 


36-9 


5.17 „ 


8 „ 


37-2 


6.15 „ 


9 „ 


374 


9.11 „ 


12 


37-6 


12.07 p.m. 


3 p.m. 


37 4 


3.03 „ 


6 „ 


37-5 


5.59 „ 


9 „ 


37-6 


8-54 „ 


12 


37-2 


2.45 a.m. 


6 a.m. 


36-9 


3.44 „ 


7 „ 


36-9 


4.42 „ 


8 „ 


371 


5.40 „ 


9 „ 


373 


8.35 „ 


12 


37-3 


11.30 „ 


3 p.m. 


37 3 


2.25 p.m. 


6 „ 


37-2 


4.20 „ 


9 „ 


37-4 


8.15 „ 


12 


375 



Mouth. 



37-0 
37-1 
365 

36 8 
37-1 
37-5 
37-6 
37-2 

37 3 
369 
367 
363 
36-4 
363 
37-0 
36-8 
36-7 
36-4 
36-5 
36-4 
362 
36-0 
36 3 
36-8 
36-4 
36-5 
369 
36-8 
36 3 
36-2 
36-5 
37-0 
363 
36-7 
368 
366 
36-2 
36-0 
36-1 
363 
367 
366 
367 
36-7 
36-6 
363 
36-4 

36 5 
36-7 
37-0 
36-8 
37-0 

37 1 
36 5 
36-7 
36 3 
363 
36-5 
36-8 
36-2 
365 
36-8 
36-7 
36-6 



Axilla. 


Air. 


36 '8 


28° C. 


366 




366 




36-8 


25 


366 


25 


367 


28 


37-1 


30 


36-3 


26 


36'8 


24 


367 


22 


36 3 


25 


36-1 


20 


36 5 


20 


36-0 


22 


36-1 


21 


36-2 


22 


366 


19 


36-1 


18 


361 


19 


36-0 


25 


361 




361 


19 


35-7 


19 


36-3 


20 


36-1 


19 


363 


19 


36-1 


16 


36-2 


16 


35-8 


23 


36-1 


19 


360 


19 


364 


20 


36-1 


11 


363 


10 


36-5 


10 


36-2 


20 


35-9 


21 


36-1 




361 


20 


36-0 


18 


35-8 


20 


35-6 


10 


36 3 


11 


362 


10 


36-2 


21 


358 


20 


36-4 


22 


365 


20 


36 


18 


36-5 


23 


36-6 


19 


36-6 


18 


366 


17 


36-2 


21 


361 


22 


36 3 


21 


36-2 


19 


359 


20 


35-9 


18 


35-8 


17 


363 


17 


36-1 


16 


362 


21 


360 


24 



Remarks. 



In laboratory ; packing books. 
In train, 15 minutes after leaving Ithaca. 
„ in bed ; feeling cool. 

After dressing in train. 

In New York City ; half hour after breakfast 

Long. 74° W. (about) ; after walk. 

In steamer, about 20 minutes after sailing. 

Walking deck ; before dinner. 

>> >> 

In state-room, before retiring. 
In bed ; open port-hole ; cool. 

)> )> 

After dressing. 
On deck ; after breakfast. 
Long. 67° 21' W. Reading on deck. 
On deck ; walking and reading. 

„ seated. 

In saloon after seat on deck. 
Before retiring ; room hot and stuffy. 
In bed ; after short sleep. 

„ after sound sleep ; room cool. 
Short walk on deck. 
After breakfast. 
Long. 59° 12' W. 
Seated on deck reading. 
In saloon, after walk on deck. 

„ reading. 

Before retiring. 
In bed ; port-hole open. 
After dressing. 

After breakfast and short walk. 
Long. 50° 38' W. Fog. 
On deck ; fog ; feel chilled. 

„ walking; foggy. 

In saloon ; reading. 
Before retiring ; feeling very cold. 
In bed. 

After dressing. 

Alter breakfast ; cold north wind. 

Long. 42° 47' W. Reading ; feeling cold. 

Playing deck billiards. 

On deck walking ; cold wind. 

In saloon since dinner, reading. 

Before retiring ; saloon had been stuffy. 

In bed ; port-hole closed ; room stuffy. 

„ „ open. 

After dressing and walk on deck. 
After breakfast. 

Long. 34° 32' W. Deck billiards. 
On deck reading ; clear ; sunshine. 
In saloon after walk. 

„ reading. 
Before retiring. 
In bed ; slept well. 

»> 
After dressing. 
After breakfast. 
Long. 25° 17' W. Fog. 
On deck ; playing quoits. 



In saloon ; reading. 



after concert in saloon. 



DAILY ROUTINE AND BODY TEMPERATURE. 



241 



TABLE I. — continued. 





Ithaca 
Time. 


Local 
Time. 


Rectum. 


Mouth. 


Axilla. 


Air. 


Remarks. 


9091 
















July 3 


2.05 a.m. 


6 a.m. 


37-0 


36-3 


361 


22° C. 


In bed. 




4.01 „ 


8 „ 


373 


36-5 


35-9 


22 


After dressing. 




4.59 „ 


9 „ 


37 5 


36-6 


36-2 


23 


After breakfast ; no walk. 




7.54 „ 


12 


37 7 


36-8 


36-4 


15 


Long. 15° 00' W. On deck ; billiards. 




10.49 „ 


3 p.m. 


376 


366 


361 


16 


On deck ; walking ; fine breeze. 




1.44 p.m. 


6 „ 


37-4 


369 


36-6 


15 


,, reading. 




4.39 „ 


9 „ 


37-4 


36-8 


36-6 


21 


In saloon ; reading. 




7.34 „ 


12 


37-2 


36-2 


360 


14 


On deck looking for Tory Island light. 


July 4 


12.28 a.m. 


4 a.m. 


36-8 


36-1 


362 




In bed ; retired at 1.30 a.m. 




2.23 „ 


7 „ 


370 


36-4 


36-4 


19 


„ no sleep since 5. 




3.21 „ 


8 „ 


373 


36-6 


35-8 


19 


After dressing and short walk. 




4.19 „ 


9 „ 


37-4 


37-1 


359 




After breakfast. 




7.14 „ 


12 


37-7 


36 9 


362 


16 


Long. 5° W. (about). Reading. 




10.11 „ 


3 p.m. 


37-8 


37-0 


36-6 


17 


Long. 4° 20' W. Standing on deck. 




1.11 p.m. 


6 „ 


37-6 


369 


36-4 




In Queen Street Railway Station, Glasgow. 




4.06 „ 


9 „ 


37-4 


371 


365 




Long. 3° 10' W. (about). Edinburgh. 




7.06 „ 


12 


371 


36-8 


365 


19 


In rooms in Musselburgh. 


July 5 


2.06 a.m. 


7 a.m. 


37-0 


36-4 


362 


21 


In bed. 




3.06 „ 


8 „ 


37-4 


369 


36-0 


20 


After dressing ; no walk. 




4.06 „ 


9 „ 


37 5 


37-1 


362 


21 


After breakfast. 




7.06 „ 


12 


37 4 


36-7 


36-4 


19 


In rooms ; reading. 




10.06 „ 


3 p.m. 


376 


37-0 


366 


20 


In Edinburgh ; after walk. 




1.06 p.m. 


6 „ 


37-5 


36-6 


36-2 


18 


After shopping. 




4.06 „ 


9 „ 


373 


36-9 


36-4 


21 


In rooms ; reading. 




7.06 „ 


12 


37-2 


36 9 


36-5 


19 


Before retiring. 


July 6 




No o 


bservatio 


up made. 








July 7 


11.06 p.m. 


4 a.m. 


36-9 


36-1 


360 




In bed ; slept since 12 30. 




1.06 a.m. 


6 „ 


36 8 


362 


36-2 


18 


» » . 5 - 




2.06 „ 


7 >, 


37-0 


362 


36-1 


17 


„ awake since 6. 




3.06 „ 


8 „ 


373 


363 


35-9 


18 


After dressing. 




4.06 „ 


9 „ 


37-4 


37-0 


36-0 


19 


After breakfast. 




7.06 „ 


12 


375 


37-1 


36-4 


21 


Packing books. 




10.06 „ 


3 p.m. 


376 


37-0 


366 


18 


" " 




1.06 p.m. 


6 „ 


37-4 


369 


36 4 


17 


After short walk. 




4.06 „ 


9 „ 


37-3 


36-8 


36-2 


16 


)! » 


July 8 


7.06 „ 


12 


37 2 


36-8 


36-4 


20 


Reading before retiring. 




12.06 a.m. 


5 a.m. 


366 


35-9 


35 8 


18 


In bed. 




2.06 „ 


7 » 


36 8 


36-2 


361 


19 


„ light sleep since 5. 




3.06 „ 


8 „ 


371 


36-7 


361 


19 


After dressing. 




4.06 „ 


9 „ 


373 


371 


36-0 


20 


After breakfast ; cold. 




7.06 „ 


12 


375 


36-8 


36-3 


19 


Packing books. 




10.06 „ 


3 p.m. 


38-0 


37 


36-4 


21 


" " 




1.06 p.m. 


6 „ 


37-6 


36-8 


363 


22 


Writing at table. 




4.06 „ 


9 „ 


37-4 


366 


361 


18 


After seat in garden. 




7.06 „ 


12 


373 


36-6 


36-0 


19 


Before retiring ; packing books. 


July 9 


1.06 a.m. 


6 a.m. 


36-9 


363 


36-0 


20 


In bed ; sound asleep since 1 a.m. 




2.06 „ 


7 „ 


37 


36 3 


36 3 


20 


„ no sleep since 6. 




3.06 „ 


8 „ 


372 


364 


36-0 


19 


After dressing. 




4.06 „ 


9 „ 


37-6 


36-9 


36-1 


19 


Reading newspaper. 




7.06 „ 


12 


38-0 


37-2 


363 


21 


Packing trunks. 




10.06 „ 


3 p.m. 


377 


37-1 


366 


18 


After short walk. 




1.06 p.m. 


6 „ 


376 


37-1 


366 


17 


" " j* 




4.06 „ 


9 „ 


37-6 


369 


36-3 


21 


Reading ; room stuffy. 




7.06 „ 


12 


37 3 


36-6 


360 


20 


Before retiring ; packing trunks. 


July 10 


11.06 „ 


4 a.m. 


36-8 


36-4 


363 




Asleep since 12.30. 




1.06 p.m. 


6 „ 


368 


36-4 


36 3 


18 


4.30. 




2.06 „ 


7 „ 


369 


36-5 


364 


19 


Just before getting up. 




3.06 „ 


8 ,. 


37-2 


367 


36-0 


20 


After dressing. 




4.06 „ 


9 „ 


375 


36-8 


362 


21 


After breakfast and short walk. 




7.06 „ 


12 


37-5 


36-7 


36-1 


23 


Working in house. 




10.06 „ 


3 p.m. 
6 „ 


37 3 


368 


36-4 


22 


Reading for half hour. 




1.06 p.m. 


37-6 


369 


365 


24 


After walk on beach. 




4.06 „ 


9 „ 


37-7 


37-0 


36-4 


26 


Working at table. 




7.06 „ 


12 


37 3 


367 


36-1 


24 


Before retiring. 



242 DR SUTHERLAND SIMPSON ON 

After arriving in Edinburgh the observations were continued, with one day's 
interruption (July 6), till July 10, the daily routine being not much different from 
the same in Ithaca. From July 17 till August 10 the time was spent in the 
Orkney Islands, somewhat over 200 miles farther north than Edinburgh, but practically 
in the same longitude, and here another series was taken from August 3 to 9 
inclusive. 

On August 5 I walked around one of the islands, a distance of about twelve miles, 
over rough ground, temperature records being taken at intervals of one and a half hours, 
and a glance at the chart (fig. 3) will show the effect on the body temperature. On 
August 8 I remained in bed the whole day, and abstained from food entirely from 
9 p.m. August 7 till 8 a.m. August 9, at the same time making an effort to restrain 
muscular action as far as possible. As these records are important for comparison with 
others taken subsequently under somewhat similar circumstances in Winnipeg and in 
Ithaca, they are given below together with the room temperature, which was practically 
the same as that of the outside air, since all the windows of the sleeping-room were 
open. Readings were taken every hour. 

The journey westward was begun on August 14, when the train left Edinburgh 
for Glasgow at 6 a.m. The steamer (s.s. Ionian, Allan Line) sailed from Glasgow 
about 10 a.m. and from Greenock about 2 p.m. the same day. The weather was fine, 
and the voyage uneventful until we ran into fog and ice off the Newfoundland coast, 
when we were compelled to go dead slow or to drift for almost two days (August 20 
and 21). On the evening of the 21st we entered the Straits of Belle Isle, and from 
that time onwards it was full speed ahead until Quebec was reached about 9 p.m. on 
August 23. That night was spent on board ship moored at the quay, and at 7.30 a.m. 
we left Quebec and steamed up the St Lawrence to Montreal, where we disembarked 
about 7 p.m. on August 24. That night and the following day was spent in Montreal, 
and at 10.30 p.m. on August 25 the train to Winnipeg was taken, where it arrived at 
9.20 p.m. on August 27. 

During the whole journey from Glasgow to Winnipeg the observations were kept 
up without interruption on steamship and train until noon August 30, three days 
after arrival in Winnipeg. 

On August 29, the second day in Winnipeg, I repeated the routine of August 8 
in the Orkneys, remaining in bed the whole day without food and voluntarily inhibiting 
muscular movement as far as possible. The light was partly obscured by the close 
proximity of a brick wall to the single window of the room I occupied, and this, 
together with the fact that no person visited me, was particularly favourable for my 
purpose, which was to secure as completely as possible bodily and mental quiescence. 
The rectal, mouth, and axillary temperatures were recorded hourly together with the 
temperature of the room (see Table II.). 

On the return journey from Winnipeg to Ithaca several stops were made on the 
way and the observations were discontinued until September 10, the day after arriving 



DAILY ROUTINE AND BODY TEMPERATURE. 



243 





Air. 


00 


OS 


o 


o 


o 


- 


i-H 


l-H 


CM 


CM 


i-H 


o 


o 


o 


o 


o 


OS 


OS 


OS 




esq 


-. 


rl 


o 


I-H 




CM 


00 


CO 


CM 


,-H 


oo 


•* 


00 


00 


CM 


CM 


CM 


r~i 


Axilla. 


CO 


CD 


CD 


CD 


CO 




CO 


CO 


CO 


CD 


CO 


CO 


CD 


CO 


CO 


CO 


CO 


CO 


CO 


of 
o 

03 








CO 


CO 




CO 


CO 


co 


00 


CO 


00 


00 


00 


00 


CO 


oo 


CO 


CO 




CM 


l-H 


CM 


-H 


-* 




oo 


■* 


CM 


t- 


T* 


■* 


Tft 


00 


Tf 


CO 


00 


rt 


© 




Mouth. 


CO 


CD 


CD 


CO 


CO 




CO 


CD 


CO 


CD 


CO 


CO 


CD 


CD 


CO 


CO 


CO 


CD 


CD 


1— 1 

C0~ 


00 


CO 


CO 


co 


CO 






CO 


CO 


CO 


CO 


00 


00 


00 


00 


oo 


00 


CO 


CO 






lO 






io 


io 


io 








iO 




IO 


IO 


IO 








iO 


■h 


Rectum. 


00 


00 


a> 


00 


o 


o 


o 


l-H 


i-H 


CO 


l-H 


l-H 


o 


CM 


CM 


CM 


^ 


o 


OS 


CD 


CO 


CO 


CO 


ir- 


I- 


1 ■. 


t- 


t- 


t- 


£- 


t- 


ir- 


1^ 


t- 


I- 


ir- 


ir- 


CD 


o 

U 

o 




CO 


CO 


CO 


CO 


CO 


00 


CO 


00 


co 


00 


00 


00 


es 


CO 


00 


00 


es 


es 


CO 




a 














a 
























* 




o3 














ft 




























CD 


r- 


OO 


OS 


o 
+- 


l-H 


CM 

l-H 


rH 


CM 


CO 


■* 


iO 


CD 


t- 


00 


OS 


o 

1 — 1 


i-H 


CM 




Air 


CD 


r- 






00 


o 


■* 




. 


OS 


CM 


o 


-* 


00 


CM 


l-H 


o 


o 


00 
















l-H 


l-H 






l-H 


CM 


CM 


rt 


M 


1-1 


l—i 


r-^ 


1-1 




Axilla. 


CO 


CO 


CO 


CO 


CO 


CM 


CM 


-* 


rii 


CM 


CM 


00 


00 


CO 


CM 


CM 


,-H 


o 


o 


of 
o 
o3 
,a 


CD 


CD 


CD 


CD 


CO 


CD 


CO 


CO 


CO 


CO 


CD 


CO 


CD 


CO 


CD 


CD 


CO 


CO 


CD 




CO 


CO 


CO 


CO 


CO 


CO 


CO 


00 


CO 


CO 


CO 


oo 


CO 


CO 


00 
CM 


CO 
CM 


CO 

o 


00 

o 


00 

© 


Mouth. 


CO 


CO 


CO 


■^ 


CO 


■* 


CM 


CO 


■* 


tH 


T* 


•G 


00 


CM 


i— i 


CD 


CD 


CD 


CD 


CD 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


CD 


CD 


CD 


CO 


CD 


CD 


CD 


co" 
CM 
u 

0) 




CO 


CO 


CO 


CO 


CO 


CO 


oo 


00 


CO 


CO 


CO 


00 


00 


00 


00 


oo 


CO 


CO 


00 






iO 








>o 




iO 




iO 




IO 






iO 








IO 


Rectum. 


O 


o 


i — i 


T-H 


o 


o 


l-H 


o 


o 


o 


OS 


OS 


o 


O 


OS 


OS 


00 


00 


t- 


HO 


t- 


l— 


£- 


r- 


t- 


ir- 


i^ 


t- 


Ir- 


t- 


CD 


CO 


t- 


ir- 


CO 


CO 


CD 


CO 


CO 


a 

as 

ft 

IP 




CO 


CO 


CO 


CO 


co 


CO 


00 


00 


co 


co 


CO 


00 


00 


co 


oo 


00 


CO 


CO 


00 




a 














a 
























CO 




03 














ft 
























* 




CO 


I- 


00 


OS 


o 

1 — 1 


~ 


CM 


l-H 


CM 


00 


Tf 


lO 


CD 


t- 


00 


OS 


o 


l-H 
i-H 


CM 

l-H 




Air. 


CD 


r- 








oo 


OS 




o 








,_, 


CM 




o 






OS 




i— 1 










l-H 


l-H 












CM 


CM 




CM 






1 — 1 












































Axilla. 


CO 


<M 


CM 


o 


CM 


"* 


lO 


CD 


iO 


Tf 


■tf 


tH 


r- 


t- 


IO 


CD 


CD 


CO 


00 




CO 


CD 


CD 


CD 


CO 


CD 


CD 


CD 


CD 


CD 


CO 


CO 


CD 


co 


CD 


CD 


co 


CD 


CO 


o3 
o 
03 




CO 


CO 


CO 


co 


CO 


CO 


CO 


CO 


00 


00 


00 


CO 


CO 


CO 


00 


CO 


oo 


CO 


CO 


Mouth. 


CO 


CO 


■* 


io 


CO 


CO 


CO 


CD 


OS 


IO 


CO 


iO 


00 


OS 


CD 


IT- 


t- 


Tt< 


CO 




CO 


CO 


CD 


CD 


CO 


CO 


CO 


CD 


CD 


CO 


CD 


CO 


CD 


CD 


CO 


CD 


CD 


CO 


CO 


of 

i— 1 




CO 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


00 


CO 


00 


00 


CO 


00 


CO 


CO 


00 


co 


CO 






iO 




•o 






1Q 




IO 


m 




o 




IO 






IO 


lO 




f* 


Rectum. 


as 


OS 


O 


1— 1 


(M 


CM 


o 


l—f 


1 — 1 


o 


i-h 


l-H 


o 


o 


l-H 


o 


o 


o 


o 


,0 

8 
ft 


CD 


CD 


r- 


r- 


I-- 


t- 


IT- 


t~ 


t- 


t- 


r- 


l> 


r- 


t- 


r- 


ir- 


IT- 


ir- 


ir- 




CO 


CO 


CO 


CO 


CO 


CO 


CO 


00 


00 


00 


00 


00 


00 


00 


00 


es 


CO 


es 


es 




a 














a 
























01 




03 














ft 
























C/2 




CD 


1— 


00 


OS 


o 

1 — 1 

-t- 


- 


CM 


i-H 


CM 


CO 


1* 


IO 


CD 


t- 


00 


OS 


o 


1 — 1 


CM 

l-H 




Air. 


CM 


CM 


,_, 


CO 


«o 






CO 






1- 






00 






00 


t- 


CD 


5C 


I— t 


1-1 


"""' 


1-1 


l-H 
























l-H 


1— t 


1 — 1 




-H 


o 


l-H 


1— 1 


l-H 


o 


o 


CM 


o 


CM 


CM 


CM 


l-H 


00 


CM 


l-H 


T+l 


© 


H 


0) 

ft 

s 


Axilla. 


CD 


CD 


CD 


CO 


CO 


CD 


CO 


CD 


CD 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


CD 




CO 


CO 


CO 


CO 


CO 


CO 


CO 


00 


CO 


CO 


00 


00 


CO 


00 


00 


oo 


CO 


CO 


CO 




,_, 


CM 


T-H 


CO 


CM 


CO 


l-H 


CM 


,_ 


i-H 


CM 


-* 


oo 


iO 


tH 


CO 


•o 


r—i 


CM 


of 

CM 


Mouth. 


CD 


CD 


CD 


CD 


CD 


CO 


CO 


CO 


CD 


CD 


CO 


CO 


CO 


CO 


CD 


CO 


CD 


CO 


CO 




CO 


CO 


CO 


CO 


CO 


CO 


CO 


00 


CO 


CO 


00 


CO 


00 


CO 


00 


CO 


CO 


CO 


00 






























IO 




IO 






iO 


Rectum. 


00 


00 


OS 


i-H 


l-H 


o 


o 


o 


OS 


OS 


OS 


OS 


o 


o 


o 


OS 


OS 


OS 


oo 


as 


CO 


CD 


CO 


r- 


1^ 


t- 


t- 


I- 


CO 


CD 


CO 


CD 


I— 


t- 


t- 


CD 


CD 


CO 


CD 


3 
< 




CO 


CO 


CO 


CO 


CO 


co 


00 


CO 


00 


00 


00 


CO 


CO 


00 


co 


CO 


CO 


CO 


CO 




a 














a 




























cS 














ft 




























CD 


r- 


00 


OS 


o 


1 — 1 

l-H 


CM 

1—t 


1—1 


CM 


00 


■* 


iO 


CO 


t- 


00 


OS 


o 

l-H 


l-H 


CM 


to 

B 


Air. 


CM 


-* 


-* 


IO 


CD 


I— 


t» 


CO 


CO 


r- 


CO 


CD 


CD 


lO 


IQ 


■<* 


■* 


*# 


CO 




CO 


CM 


r— 1 


iO 


IO 


oo 


oo 


-* 


00 


CM 


CO 


■* 


CO 


iO 


O 


r—i 


OS 


l-H 


00 


eg 


Axilla. 


CD 


CO 


CD 


CO 


CD 


CD 


CO 


CD 


CD 


CD 


CD 


CO 


CD 


CO 


CO 


CD 


IO 


CD 


iO 


as 

l-H 




CO 


CO 


CO 


CO 


CO 


CO 


00 


00 


CO 


CO 


CO 


CO 


CO 


00 


00 


CO 


CO 


00 


CO 




-p 


CM 


CO 


JC- 


■>* 


CO 


iO 


CM 


rH 


CO 


IO 


CM 


■* 


Ttl 


CM 


o 


© 


i-H 


00 


3 
in 

O 


Mouth. 


CD 


CO 


CO 


CO 


CO 


CD 


CD 


CO 


CD 


CD 


CD 


CO 


CO 


CD 


CO 


CO 


CD 


CD 


iO 




CO 


CO 


CO 


CO 


CO 


CO 


00 


00 


CO 


CO 


CO 


CO 


00 


CO 


CO 


CO 


00 


00 


CO 








io 








iO 


»o 




iO 






iO 


iO 














Rectum. 


OS 


00 


os 


CM 


l-H 


o 


o 


o 


l-H 


o 


o 


^H 


o 


OS 


00 


00 


00 


00 


l- 


00 


CO 


CO 


CO 


£- 


r- 


r- 


£- 


I> 


t- 


t- 


t- 


t- 


t- 


CD 


CO 


CO 


CO 


CD 


CD 


*3 
QD 

So 

b 




CO 


CO 


CO 


CO 


00 


CO 


00 


co 


00 


00 


00 


00 


00 


CO 


CO 


CO 


CO 


00 


CO 




a 














a 
































































«1 




03 














ft 


























lO 


I-- 


00 


OS 


o 


i-H 


CM 


l-H 


CM 


00 


tH 


iO 


CD 


t- 


00 


OS 


o 


l-H 


CM 
















1-1 


T— 




















1—1 


l-H 


l-H 



13 
B 
03 



=4H Ol 
^ ,H 

03 S 

<u +r2 

U 03 

r> in 

i— 1 as 

00 ft 

u a 

°£ 
og 

« a 
o 5 



3 ti 

^ to 

o3 X 



■^ a 

2^ 

Pi =<H 

U2 o 

^^ 

rrl CS 

oi S ft 

S "CO 

0) ?_, 
+3 ^h (M 

a a « 

o B <u 
03 r ^j 

h a <o 

a a o 

ft 03 OS 
O OS g 
OS Ch 03 

-a os o 

^ .+* H 
**-* Hh 

a 03 o 
»^§ 

CO as -2 

u g a 

as ? as 

-g-s ^ 

a % os 

O^s - 
■-d as o 

S H as 

03rO eg 

CD CM as 
CM i-h % 



ft ft S3 
os os 

a a - 



03 



244 



DR SUTHERLAND SIMPSON ON 



o 


•s 


•n 


Tl 


03 


03 


03 




a 




03 

"1 


3 


43 

4^ 


p4 


09 

43 




ri 


be 


CJ 


tf 


OhT) 






43 
*-> 


& 


J 

B 








O 




hH 


03 


•*"* 




03 


CO 




2 


m 


CO 


ri 


+^ 


ri 


3 
O 




o 


OJ 


kl 




■P 




S3 




>-■ 








oj 








tr 


-O 


-3 
H 


rr 


00 
ri 






tea 


-d 


'T 


>» 


Pj 




cS 


p 


CO 


CD 


re 


► 


03 




> 

c3 












4^ 




P 


Tr. 


'■+-> 


01 


p 


09 




tr 


ri 


B 


ri 


-3 


O 








03 

S 


o 
-^ 

a 


o 
O 


5 


P 
^2 


43 

3 




^~ 


O) 




ri 








of 


r - 


cfl 

o 
43 


5 


4S 
he 


i— i 


03 

bo 








+j 






B 




ri 
43 
40 


o 

CD 
(D 


o 
4-a 


03 
CO 

a 


O 
i— 1 


ri 
C3 




3 


4-a 


,£> 


00 

p 


03 


CD 
rQ 


ri 


03 

-• 


_ 


03 

O 


>> 


— 


z! 


a 


T3 

ri 


ri 


o 


B 

03 


o 

C3 


03 


0^ 


&4 

60 


Ml 






O 




03 


03 


B 




P 






— 


x 


43 


03 
03 
03 


<* 


03 

P 


a. 

43 


43 

73 


3 


2 
o 




03 

43 



03 

P 
O 

P 

.5 


O 

P 
O 


O 


s 


Cf-i 

-a 
B 

ri 


E 

ri 


bO 
03 
-^ 
03 


o< 


03 


O 


CO 

a 
O 




ri 


'eo 


X 


o 


T.1 


5 


09 


+^ 




B 




G> 






a. oj 


O 




-H 


>s 




o 


'?, 


r-H 


B 
o 


j= 


03 


ea 

03 


H 


— 1 

03 

43 




B 


CD 


03 








ort 








A 


ri 


+j 


G 


a 


03 


4-> 


3 


H 


ri 




3 





.~ 


4^» 

ri 


2 




03 


"° 


>• 


03 
Tl 


^3 


p 

ri 

p 


c 
s 


o 
09 
43 


a/ 


43 

2 

ri' 


43 

a 


B 

'S 


o 

ri 


,M 


.s 


o 




>3 


43 
CO 


o 




=1 


2 


to 


o 




> 


4J 

ri 


m 


01 






4-= 


43 


O 




03 


n 


01 











5 


co 

a 

s 


+3 




s 


S3 


4a 


03 




£ 


P* «o 


^> 


'■d 


n 


Oh 










ri 


ri 




2 

03 


03 


03 


so 


>* 

ri 
-a 


~i 


CO 


5 


&►, 





a 


ri 


Tj 


ri 


f£ 


03 


m 


•- 


TJ 


id 


B 


.M 


03 


~ 
























'■J 




ri 


ri 








+-> 


3 




T3 


>, 




53 


+3 


tj 


eg 


3 


cp 


C 


td 










c3 

to 






a 


SO 


d 


a 




93 


a 




B 




EC 

o 

a 


o 


ri 




^Q 


43 

3 
o 

2 

d 

j3 

O 


43 

w 

2 

p 


bO 

a 

03 


pVi 
to 


03 

t> 

03 
CO 
43 


B 
O 

d 

"g 

ri 


03 
+3 

ri 

03 

ID 

5 


0j 

a 


>> 


CO 


'3 

> 

o 


03 

43 




M 

ep 




BO 

O 

> 
03 


'3 

03 


40 

43 

to 


'te 


09 

03 
> 


03 
















C*H 






Oj 






- 




o 





8 


2 




03* 


CD 




o 




a 


- 


Cue 


s 


+3 





a 












ri 




(U 


40 








43 


C3 








e 


03 




-( 


O 




0) 


'f. 


£ 


-^ 


T3 

CD 


— 


r/J 


+a 


»j" 


a 


%-> 


► 


3 


43 


Ph 


4- 




••* 





~ 




2 


"^ 


ri 


> 

a 


~r 


o 

03 


>. 


4-> 

ri 

03 






2 


on 


5 


~h 


3 


43 


ri 


CO 
09 


iO 










03 




■a 


1 


~ 
H 


43 


ce 

5o 


V 


5 

CO 


'ri 
S 
40 


a 

ri 
X. 



liiUll 


iiP^s 


lllllll 


\^v>y>i^i*v^« 


llillll 


.-^^*-- 1 


lllllll; 


liipilr 


















Siissui 


iiiiiilis 


lllllll 


y j J 


lllllll) 




iiiilli 


liilipi' 








' 5-y 




^^i*- 




^"SBa-*--- 


|§i§SSS: 


^ A ■ 


lllllll 


i w^-n/s-5 : 5>s« 


lllllll 


^«^^ ' ■ 1 


iiiiiii 


i^pl|§* 








* t_ ~v 




2 c 




^5""'*^ 


sssiliS 


<s » ^ s^ s^ ^ ^»B 


lllllll 


v^§>^S^$;^» 


lllllll 


S <S * '& K ss i^fl 


sllsssl 


Sll^^S^si 








^^■^ 










sssssis 


li^ni 


§§§§§§§ 


*S«s:«S!SM 


llillll 


|^^^^^S« 


§ii§il§ 


UlllS^ 


X z 




-< ->% 












SSSSSSS 


ssis^i 


lllllll 


ss^^s^^ 


§|ss§ss 


^^^p^s^B 


ss^Ssss 


^^^^^^^P 




f-^ 




^-^ 




^ 






§lliss§ 


%w§k%sm 


llllllg 


"" *±xt*^ H 


llllils 


iiiinn. 


ilsipi 


P^sssal 


















SS§|§1§ 


5^§^^^^k 


lllllll 


^^^^^^^B 


iiinii 


^^^^^^i« 


silisli 


llsipl^i 










^ 








SSSSSSS 


^?: vx i*?^ 


lllllll 


r j • • 






S^SSS^S 


■5^'^X^'^^«b| 






5^ 




*t* 








liissgi 


^ipss^^ii 


sssssSS 'S 


t^^^SxssB 


illliis 


ipiilB 














•• ^ 


ill 




^^ p 




/ V 


^ 


» """■. Li 



"3 



B 
=1 



13 



ri 



_B 
T3 



S P 



O 



S '- 1 



,s§: 



w 



S T3 



itMOOOCO-*CMOOOCO-*(NOCOCO 



woanO'J'wotoo^wotflO 



DAILY ROUTINE AND BODY TEMPERATURE. 



245 



S is 



!> 



B S 




(B 




-o 


3 
so 




£ 


3 




-M 


< 


t" 


05 






r/7 


3 


a 


rC 




f^ 


o 


rt 




1—1 


o 






X 














s 

= 







a i 



NOao^wOM^^NQQO ^o 




P' 


5^5 i 


> 








~ 


ss 




i 


i 




l*" 


>tff<7 


r 
































































^ 


$fc 


> 












L- / 






y 




■ 


^P^ 






r 


^m 


I 


^ 






- 


: 




m 


■ 
















■ 






























































''••"'- 






■ 










f 


lP 














.. 




• 










^ 








'•• 


j{ 




















































^ 






























iy 




■' 








"■■■■' 


"" 


""l 


>:: 




"" 




















































































H 




















i 


^ 


. 












1 


^ 












•- 














































,<i 


























< 




' 




^ 


















J 




■ 






, 




.■- 


,-' 




1 


: 


^ 












■ 






\ 










1 






















































^ 


^ 


\N 




$& 


* 


^ 












, 










i 












, 


/ 




,. 


\ 


^ 


v 






' ■ 






















) 






\ 












\ 






^ 


\ 




















i 





a 



3 2 



3 
60 
3 
"4 



coo(o*M3a)(e*MOw 



246 DR SUTHERLAND SIMPSON ON 

in Ithaca, when they were resumed again for four consecutive days till 6 a.m. on 
September 14. On September 12 and 26 and October 3 the three twenty-four 
hour periods were spent at rest in bed, as on the former occasions in Winnipeg and 
Scotland. On September 12 and October 3 the three ordinary meals were taken, 
but on September 26 food was abstained from (see Table II.). 

IV. Results and Discussion. 

The results of this continuous series of observations from June 19 till September 14, 
with the interruptions as indicated, are presented in the accompanying temperature 
charts (figs. 1 to 5). The thick continuous line represents the rectal temperature, the 
dotted line the mouth temperature, and the thin continuous line, usually the lowest in 
position, the axillary temperature. The division of the twenty-four hours into night 
and day periods is indicated graphically by the alternate light (6 a.m. to 6 p.m.) and 
dark (6 p.m. to 6 a.m.) bands at the bottom of the chart according to Ithaca local time. 
The hatched and intermediate clear segments extending through the whole width of 
the chart shows night (6 p.m. to 6 a.m.) and day (6 a.m. to 6 p.m.) by ship and train 
local time, and the alternate light and dark bands at the top the same by Edinburgh 
time. This division is somewhat arbitrary, since from June 19 till September 21 the 
sun rises before and sets after 6 o'clock, so that the dark and hatched bands on the 
chart are more extensive than the natural night period, but the attempt to construct a 
chart in which these should coincide exactly with the dark periods for the different 
latitudes in which the observations were made was found to involve too much labour 
and was abandoned. In the Orkneys a newspaper could be read outside the house at 
10 p.m. and again at 2 am., the actual dark period at that season of the year being of 
very short duration. As it stands this feature of the chart is only meant to show in a 
graphic way the relative differences between the local time on the journey and that of 
Ithaca and Edinburgh. From June 19 to 25 it will be observed that the dark and 
hatched bands coincide, and also from September 10 to 14; but on the voyage east- 
ward the ship's local time moves forward, and on the voyage westward it moves back- 
ward in relation to Ithaca time. 

The figures at the left side of the chart represent degrees centigrade, those at the 
bottom hours of the day, and those at the top days of the month. Twelve o'clock 
noon occurs at the middle of the clear band and 12 o'clock midnight at the middle of 
the hatched band, each space between the vertical lines representing an interval of 
three hours. 

A careful scrutiny of this chart will bring to light some interesting facts, and the 
first and most obvious of these is that the body-temperature curves of no two days are 
alike. If inherent rhythmicity were a predominant feature, one would expect to find 
a closer resemblance between the curves on different days, whereas they all seem to 
be controlled to a very large extent by the conditions imposed on the body. This is 



DAILY ROUTINE AND BODY TEMPERATURE. 247 

well seen on June 24, 25, and 26, when the external temperature was high and muscular 
movement only slightly more active than on the four preceding days. It shows what 
a powerful influence even very moderate exercise has on the body temperature in hot 
weather. On the 24th at 3 p.m. the rectal temperature reached 38°'5 C, and at 7 a.m. 
the following day it was 36°'7 C, giving a range of almost 2° C, about double the 
average for the previous five days. 

After getting to sea on the homeward journey cold, foggy, depressing weather, such 
as is usually found on the banks of Newfoundland, was experienced for the first four 
days, and its influence is at once evident on the temperature curve. The maximum 
falls about l c C. between the 26th and 27th, and the diurnal range is also greatly 
reduced. The lowest point was reached at 3 a.m. on September 30, when the rectal 
thermometer stood at 36° '2. From June 30 till the end of the voyage the weather was 
clear, bright, and warm, with sunshine throughout the greater part of the day, and here 
again the effect on the body temperature is at once visible. 

The influence of severe muscular exercise with a moderate external temperature is 
well seen on August 5, when a long walk was taken over rough ground ; the rectal 
temperature was maintained over 38° C. for several hours, while the air temperature 
averaged about 1]° C, with a cool breeze blowing. Again the influence of a high air 
temperature with moderate muscular exercise is illustrated by the rectal temperature 
curves of August 22, 23, and 24. On the 22nd the ship was in the lower reaches of 
the Gulf of St Lawrence, just through the Straits of Belle Isle, and not far from ice. 
As she steamed up the Gulf and into the river the air temperature rose rapidly until 
it reached the neighbourhood of 31° C. near Montreal on the 24th. 

On the train sleep was poor, and the minimal rectal readings were high on August 
26 and 27, probably on this account, while the effect of a sound sleep following a 
period of fatigue is seen on the morning of the 28th, when at 3 o'clock the body 
temperature (rectal) sank to 36°'7 C, a point lower than had been reached on any 
previous occasion during the westward journey with one exception, 4 a.m. August 17. 
The contrast between the curves of August 5 and 29 shows the effect of muscular 
exercise, the external temperature being not much different on the two occasions. 

Proceeding now to the discussion of the main question (to throw some fresh light 
on which the present investigation was undertaken), viz. whether the temperature curve 
is controlled by purely local and external conditions, or shows evidence of an innate 
periodicity independent of these, I believe that we shall find the evidence to be mostly 
on the one side. If the temperature rhythm were fixed in the observer's body we 
should expect to find that as he proceeded eastward, gaining time daily, the morning 
(6 to 9) rise would begin later and later each successive day until the end of the journey, 
and when Edinburgh was reached, where 11 a.m. corresponds with 6 a.m. at Ithaca, this 
rise should show itself between 11 a.m. and 2 p.m. In this relation we shall consider 
the rectal temperature only, since this is influenced to a smaller degree by accidental 
external disturbances than either the buccal or axillary temperature. A careful 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 12). 38 



248 



DR SUTHERLAND SIMPSON ON 



37° -0 



37*0 



37°0 








kH 





3 6 9 12 3 6 9 

Flo. 6. — Average temperatures 
for different periods. 



examination of the chart (figs. 1 and 2) will show that no 
such delay is apparent in the diurnal curve. The sharp 
morning rise is not any less abrupt on July 5 than it is, for 
example, on June 20. # The curve appears to obey local 
time and not Ithaca time. 

In order to get rid of accidental variations which might 
affect the body temperature on any one day, the average 
curve for each of the following periods is taken : — A, six 
days in Ithaca, immediately before setting out on the east- 
ward journey, from June 19 to 25 (omitting one day — 
24th — when the rectal temperature was distinctly above 
the usual level, due to a combination of muscular exercise 
and high external temperature) ; B, the first four days 
(June 27 to 30), and C, the last four days (July 1 to 4), 
of the voyage from New York to Edinburgh ; D, five days 
immediately following arrival in Edinburgh (July 5 to 10, 
omitting 6th) ; E, five days in the Orkney Islands about 
one week before starting westward (August 3 to 13, 
omitting 5th, 8th, 10th, 11th, and 12th); F, the first six 
(August 15 to 20), and G, the last six days (August 22 
to 27, omitting 24th) of the journey from Scotland to 
Winnipeg ; H, one day and a half at Winnipeg (August 28 
till noon on 30th, omitting 29th) ; and lastly, I, the four 
days immediately following the return to Ithaca from 
Winnipeg (September 10 to 14, omitting 12th). In these 
average curves the rectal temperature only is considered. 
(See Table III.) 

From an examination of these nine mean curves 
(charts A to I, fig. 6) it will be seen that although there 
is considerable variation in their general characters, the 
most constant features, viz., the rise between 6 a.m. and 
9 a.m., and the fall from 9 p.m. to midnight, are fairly 
regular. Compare, for example, A. the Ithaca control, and 
C, the mean of the last four days of the journey eastward. 
Except for the slightly smaller range of C, due probably 
to the less active life led on board ship, they very closely 
resemble each other. The first maximum is reached at 

* On two or three occasions in the control period at Ithaca the rapid rise 
from 8 a.m. to 9 a.m. is associated with a walk up a rather steep hill after 
breakfast in hot weather and shortly before the 9 o'clock observation was 
made. This is apt to give a wrong impression when comparing the curves at 
Ithaca with those obtained subsequently. 



DAILY ROUTINE AND BODY TEMPERATURE. 



249 



noon and the second at 9 p.m., the same hours in both cases. The ascent from 7 a.m. 
to 9 a.m. is slightly less steep in C than in A, which might be explained by the 
fact that on two or three occasions in Ithaca the 9 a.m. temperature was taken very 
soon after a walk uphill in hot weather, but the descent from 9 p.m. to midnight is, 
on the other hand, somewhat more sudden. If the Ithaca temperature habit had been 
fixed in the body, there should be some evidence of delay in the rise of the curve C 
and also in its fall, but there is none. These curves A and C resemble each other more 
nearly than any other two, except probably D (Edinburgh) and G- (the last half of the 
voyage westward), where the time difference is again four to five hours. 

TABLE III. 

Average Bectal Temperature for Nine Periods as stated in the Text. 





June 19-25. 


June 27-30. 


July 1-4. 


July 5-10. 


Aug. 3-13. 


Aug. 15-20. 


Aug. 21-27. 


Aug. 28-30. 


Sept. 10-14. 


6 a.m. 


3692 


36-55 


36-95 


36-79 


3680 


3693 


3697 




36-83 


7 „ 


36-90 


36-60 


36-90 


36-94 


3687 


36-92 


37-12 


36-90 


36-90 


8 „ 


37-16 


36-80 


3722 


37-24 


3693 


37-15 


37-33 


36-95 


37-10 


9 „ 


37-45 


37-07 


37-40 


37-46 


37 35 


37-47 


3760 


37-45 


3737 


12 


3763 


37-12 


37-57 


37-58 


37 43 


37-48 


3767 


37-55 


37-70 


3 p.m. 


37-58 


36-90 


37-52 


37-64 


37-65 


37-45 


3772 


37-40 


37-66 


6 „ 


37-55 


3687 


37-42 


37-54 


37-52 


3755 


37-70 


37-40 


37-80 


9 „ 


3760 


37-07 


37-45 


37-46 


37-43 


37-47 


37-58 


37-20 


37-70 


12 


37-42 


3690 


3725 


37-26 


37-10 


37-07 


3732 


37-00 


3730 



It might be supposed that after a five weeks' residence in Scotland the body would 
gradually accommodate itself to the changed conditions, and a Scottish temperature 
rhythm be established. This would be present in the mean of the observations made 
in the Orkney Islands — curve E ; and if this type had been carried westward, then 
at Winnipeg, where the Scottish daily routine had been moved backwards by more 
than six hours, we would expect to find that the curve for Winnipeg (H) should 
show an upward tendency earlier, according to Winnipeg local time, and also decline 
earlier than the Orkney curve (E). It does appear that the fall sets in sooner, but the 
Winnipeg type represents only a single day, and some accidental variation might 
account for the early decline. The morning rise does not come in any sooner than in 
Scotland. 

In making such comparisons, however, it is essential that the conditions with regard 
to external temperature, meals, mental activity, and particularly muscular movement 
shall be as nearly as possible similar at the two places, and this will be best secured 
if the subject remains at rest in bed, and at the same time abstains from food while the 
observations are being made. The effects of external factors will then be reduced to 
a minimum, and any inherent temperature rhythm present in the body may be 
expected to show itself. Assuming that the Ithaca rhythm has been replaced by a 
Scottish rhythm in the course of a six weeks' residence in Scotland, then if the latter 
had persisted in the body to any appreciable extent one would expect to find that the 



250 DR SUTHERLAND SIMPSON ON 

records obtained at Winnipeg on August 29 — -with the subject resting in bed and fasting, 
as in the Orkneys on August 8— when plotted out should give a curve different from 
that found in Scotland : the morning rise as well as the late evening fall should appear 
some six hours earlier in the Winnipeg than in the Orkney curve. Such, however, 
is not the case. The morning ascent, to be sure, is slightly less abrupt in the Winnipeg 
curve, but this may be accounted for by the fact that here the subject was allowed 
to remain quite undisturbed, whereas on August 8 he was persistently invited to 
partake of some breakfast by kind but injudicious friends. 

There is no evidence of any persistence of the Ithaca control type of curve in 
Scotland, nor of the Scottish type in Winnipeg. In the westward as in the eastward 
journey there appears to be an immediate adjustment of the temperature rhythm to 
the changing daily routine, so that in this respect the results agree entirely with those 
obtained by Gibson. 

V. Effect of Muscular Rest and Activity on Body Temperature. 

After plotting out the curves for the rectal temperature readings taken on August 8 
and 29, I was struck with the marked effect which complete rest in bed has on the 
diurnal body -temperature variation, and to test this further, I made experiments 
under nearly similar conditions on myself on three occasions after returning to Ithaca 
(September 12 and 13, 26 and 27, and October 3 and 4) and twice on two other 
individuals (November 27, 28, and 29, and January 29, 30, and 31). 

On September 12 I stayed in bed all day, i.e. from 12.15 a.m. till 7 a.m. on the 13th, 
and, as on August 8 and 29, endeavoured to make as few movements as possible, but 
on this occasion I had the usual three meals, taken in bed in the recumbent position — 
breakfast at 9.30 a.m., lunch at 1.30 p.m., and dinner at 6.30 p.m. The temperature 
of the room rose steadily from about 16° C. at 6 a.m. to 22° C. at 7 p.m., and then fell 
to 19° C. at midnight. Such changes in the outside temperature, however, in the case 
of a person in bed and covered with blankets, are of little importance, except as affecting 
the temperature of the mouth (fig. 9). 

On September 26 the same experiment was repeated, with this difference, that no 
food was taken from 9.30 p.m. on the 25th till 8 a.m. on the 27th. The bed on 
this occasion was outside on the verandah, and the air-temperature range was from 
6° C. at 6 a.m. to 22° C. at 4 p.m., after which it fell regularly till midnight, when 
the thermometer stood at 8° C (fig. 10). 

On the following Sunday (October 3) I again remained in bed on the verandah, 
but had breakfast at 9 a.m., lunch at 2.30 p.m., and dinner at 6.30 p.m. The air 
temperature was about 10° C, and varied little throughout the day (fig. 11). 

Wishing to ascertain how this modification of the daily routine would affect the 
diurnal temperature variation in other individuals, I prevailed on Miss A. B. and 
Mr X. Y., workers in the laboratory, to make similar experiments on themselves, and 



DAILY ROUTINE AND BODY TEMPERATURE. 



251 



38°-0 


Or] 


cne 


yi 


sla 


ids 


, / 


LUg 


usf 


St 


h 
























9th 






















































































































4 


































































































































j 


* 




























— o 


















































/ 














































































/ 


































6 
































f T T TSi-- 


«a 







































5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 C 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 

A.M. P.M. A.M. P.M. 

Fig. 7. — Effect of muscular inactivity and fasting on the rectal temperature. 
Winnipeg, August 29th 30th 



37°0 



M^_ 


~f_ 


/* <k N, ,, ,. it -J>~_. ) 


t y > 'Shhm^TShhh,-.-.-h« / 



56789 10 11 12 12 3 456789 10 11 12 6789 10 11 12 123456789 10 11 12 

A.M. P.M. A.M. P.M. 

Fig. 8. — Effect of muscular inactivity and fasting on the rectal temperature. 
Ithaca, N.Y., September 12th 13th 14th 



38° -0 



4 

2 

37°-0 



:^^:=^=^: 



5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 
A.M. P.M. 



789 10 11 12 1234567 
A.M. P.M. 



10 11 12 6 



Fig. 9. — Effect of muscular inactivity on rectal temperature. 



2 


It 


ha 


sa, 


N. 


Y. 


S 


;pt 


jm 


jer 


26* 


h 
























2 


7tl 


l 








A 


i, 




















2S 


th 




























































/ 


\ 








































































J 


\ 




/ 








\ 


^ 


<""* 


V 






























































1 


< 




s 


i 














V 


>— « 




























































/ 


















































,—< 


»— i 


, — i 


1 r-A 


\. 






















r— i 


f-f 


/ 






































37 '0 
8 




















s 


►-< 


r- 4 














, r 


/" 






































* 








































S 


r 















































7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 6 

A.M. P.M. A.M. P.M. 

Fig. 10. — Effect of muscular inactivity ami fasting on rectal temperature. 



2 

38°"0 


Itl 


lac 


»,] 


ST."! 


f.i 


Oc 


tol 


er 


3r< 


1 




























itl 


































5 


th 


g 






















































































6 






















































































4 






















































































2 






















































































37°0 






















































































8 






































' > 


► _ 










































*, 




6 
















































































I 


> 



5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 (I 

A.M. p.M. A.M. P.M. 

Fig. 11.— Effect of muscular inactivity on rectal temperature. 



252 DR SUTHERLAND SIMPSON ON 

I would take this opportunity of expressing ray indebtedness to both for their ready 
consent to pass through this, to healthy young people, somewhat trying ordeal. 

Miss A. B. (age 27, weight 116 lbs., height 5 ft. 1\ ins., and of slender build) 
made observations at irregular hours on November 27 and 29, while engaged in her 
ordinary work at the University, to obtain curves on days of activity for comparison, 
and on the 28th took readings every hour, keeping bed from 10.15 p.m. on the 27th 
till 6 a.m. on the 29th, at the same time abstaining from food, " except a glass of 
milk and some crackers between six and seven in the evening." When examining her 
figures I found that twice there were somewhat sudden elevations in the temperature 
(at 3 p.m. and 6 p.m., fig. 12), which then quickly dropped again, and on writing to 
her to find out the cause of these, her reply was: "... I read some, slightly propped up 
in bed. The rises in temperature came at the periods when I was disturbed by callers. 
I tried to keep my room at the same temperature as I had it at night — heat off and 
window open. I remember that during the afternoon I was very restless, even 
when alone." It is quite evident to me that she had difficulty in restraining her 
muscular movements to the same extent that I did, and this is shown in her 
temperature chart for that day (fig. 12). 

Mr X. Y. (age 27, weight 134 lbs., height 5 ft. 8^ ins., and sparely built) also continued 
his experiment for three days. On January 29 and 31, 1910, in his rooms and while 
at work in the laboratory, he made observations every two hours, and on the 30th 
every hour from 5 a.m. till 10 p.m. Like Miss A. B., he kept bed from 10 p.m. on 
the 29th till 6 a.m. on the 31st, but he had three meals at his usual hours — breakfast 
at 8 a.m., dinner at 1 p.m., and supper at 6.30 p.m. Besides, he was reading most 
of the time, and this implies a fair amount of muscular exertion, since in the recumbent 
posture a book is not easily held in front of one in such a position that it can be read. 
In this experiment, although the curve is distinctly modified when compared with that 
of the day before or the day after (fig. 13), the temperature range is considerably 
greater than in either the case of Miss A. B. or that of the writer. It may be 
mentioned in passing that neither of these (A. B. nor X. Y.) was a good subject for 
this experiment, since both are naturally energetic and somewhat restless individuals. 

In the literature of this subject which I have consulted several temperature charts 
are presented from healthy individuals " resting in bed " during the day, and most show 
a very distinct diurnal variation. For example, in Pembrey's article on "Animal 
Heat" in Schafer's Text-book of Physiology (vol. i. p. 800), a curve, copied from 
Ringer and Stuart,""' is given which " shows the daily fluctuations of temperature in a 
boy 12 years old; the thermometer — a non-registering one — was kept in the closed 
axilla throughout the time, and the readings were taken every hour. The boy was in 
good health, and was kept in bed during the observations," which extended over fifty 
hours. The range for each of the twenty-four hour periods is almost 2° C. (from about 
36° to about 38°). I cannot help thinking that this must have been a particularly 
* Ringer and Stuart, Proc. Roy. Soc, Lond., vol. xxvi., 1877, p. 187. 



DAILY ROUTINE AND BODY TEMPERATURE. 



253 













































































,'" 




_^ 




/ 




' 




"1 




















































J 




,r 


































































\ 


































i 



o 









3 H 


































































































































































































































































+3 


















CO 
























































































t 


































i 


















a 
















s 


-° 














J 
















y 


<* 


































\ 


















\ 


\ 


















I 


















\ 
























































































CO 












\ 




















v 


















N 


i 


















\ 


















> 
















• 
















* 
















• 
• 
















































































o 


















OJ 


















^r 




































>> 


















3 


















1-5 


















: 


















kH 


















& 


















o 


















4-3 








































254 DR SUTHERLAND SIMPSON ON 

restless boy. The fact that the temperature was taken in the axilla need not 
necessarily vitiate the results when the subject remained in bed all the time, 
although it would have been more satisfactory if the records had been obtained from 
the rectum. 

Again, in a recent paper by Bardswell and Chapman,* a curve is given showing 
the average daily variation in the rectal temperature of nine healthy individuals (seven 
men and two women, all between the ages of 20 and 35 years) who were kept in bed 
during the period of observation, which varied in different cases from twenty-four to 
forty-eight consecutive hours. The ordinary meals were taken at 8.30 a.m., 1 p.m., 
and 7 p.m. During the night the individuals were aroused sufficiently to allow of the 
temperature being recorded, and sleep was very little interfered with. Readings were 
taken every two hours throughout the twenty-four. The minimal temperature occurred 
between midnight and 4 a.m., the average for the nine individuals being 3 a.m. The 
average minimal temperature observed was 36° C, the lowest being 35 0, 8 and the highest 
36°'3. " From the minimal level the temperature rises gradually until the period of 
sleep is completed (the average waking temperature being about 36°*4 C), and there- 
after it rises rapidly until at 10 a.m. the temperature is 36°*9 C. From 11 a.m. till 
6 p.m. the temperature remains practically steady between 36°'9 and 37°"1, the highest 
point usually being reached between 4 and 6 p.m. After 6 p.m. the temperature begins 
to fall again, but only gradually. When the period of sleep is reached the fall continues 
rapidly until at about 2 a.m. the temperature curve is again at its minimal level." In 
this average curve the range from 6 a.m. till midnight, which corresponds with the 
waking periods in our experiments, is about 0°'8 C. This is greater than is found in 
any of our curves except that of Mr X. Y. 

Johansson,! on the other hand, by enforcing muscular rest and at the same time 
fasting, was able to reduce the temperature range to 0°"4 C. in the twenty-four hours 
including the sleeping period. Hormann J also found in the case of an insane woman 
who remained in bed and took no food for three days that the diurnal fluctuations in 
the vaginal temperature were almost abolished. 

" Rest in bed " is, of course, a relative phrase, and may have a different meaning for 
different individuals, but to remain in bed during the waking hours and voluntarily 
inhibit all muscular movements except when these are absolutely necessary, and at the 
same time prevent one's self from falling asleep, is a task not easily accomplished. As 
a matter of fact, this experiment is not likely to be carried out properly by any subject 
who is not himself particularly interested in it. 

Muscular contraction is the most potent factor in elevating the body temperature, 
and sleep in lowering it. Benedict § found that a change in body-position from sitting 

* Bardswell and Chapman, Brit. Med. Jour., May 1.3, 1911, p. 1107. 
t Johansson, Skandinavisches Archiv fiLr Physiologie, viii., 1898, p. 85. 
I Hormann, Zeitschrift fur Biologie, xxxvi., 1898, p. 319. 
§ P.enedict, Amer. Jour, of Physiology, vol. xi., 1904, p. 167. 



DAILY ROUTINE AND BODY TEMPERATURE. 255 

to standing was sufficient to produce almost immediately an appreciable rise in the 
rectal temperature, and vice versa when the change is from standing to sitting. This 
rise he attributes to the muscular work involved in maintaining the body in the erect 
attitude. However, there appears to be considerable variation amongst different 
individuals in the temperature response to muscular exercise. In the observations of 
Leonard Hill and Martin Flack,* made on healthy athletes in a three-mile race, they 
found in one case at the end of the race the extraordinary high figure of 105° F. 
(40°*56 C.) for the rectal temperature. As a rule they found that "the longer the 
effort the higher the body temperature rose," within the limits of the race of course. 
"Thus W. V. F. was 101° 1 F. after i mile, 102°-0 after | mile; H. 102°'8 after 
1 mile, and 103°'6 after 3 miles; H. P. 102°-8 after 1 mile, 103°-8, and 105°-0 after 
3 miles. The temperature did not rise in all individuals. Thus J. F. P. showed no 
rise after 1 mile, or after seven laps of the three-mile race." From systolic blood-pressures 
taken at the same time they attribute the high temperatures, in part at least, to the 
influence of cutaneous vaso-constriction. 

For the same individual, however, the same amount of exercise taken within the 
same time limits will produce roughly the same rise in rectal temperature. Thus 
Bardswell and Chapman,' 1 ' from experiments conducted on themselves and other 
healthy persons, found that the rise of temperature produced by definite amounts of 
exercise were so consistent and constant that they were able at will to raise their 
temperatures from the normal to any point up to 103° F. (39°"5 C.) simply by "pre- 
scribing to ourselves varying degrees of muscular effort. By dint of constant observa- 
tion we were able to guess to within a point or two what our temperature would be at 
any time of the day and after any kind of exercise. " 

The effect of food on raising and maintaining the body temperature can be seen 
by a glance at the five curves of the writer, taken on the resting days (figs. 7-11). 
Three meals were taken on two of these days, viz. September 12 and October 3, 
and on these days the rectal temperature is distinctly higher, and on October 3 also 
less regular, than on the three fasting days. There does not appear to be any im- 
mediate effect following the meal, but the general level of the curve is higher by about 
three-tenths of a degree centigrade after the first meal on September 12 than on the 
three days when no food was taken, and this difference is maintained throughout the 
day. On the fasting days the decline also sets in earlier. 

In the case of Miss A. B. the curve is somewhat distorted (fig. 12) by the 

comparatively high figures at 3 p.m. and 6 p.m., but that these are due to 

accidental disturbances is indicated by the low temperature before and after these 

hours. The probable cause of these disturbances she explains in her note. Apart 

from that and from the fact that the general level is somewhat higher, the curve 

does not differ materially from those of the writer, taken on the three resting and 

fasting days. 

* Hill and Flack, Jour, of Physiol, Proceedings, xxxvi., 1907, p. xi. + Bardswell and Chapman, loc. cit. 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 12). 39 



256 DR SUTHERLAND SIMPSON ON 

From the preceding remarks will be seen the importance of making full allowance 
for the elements of muscular exercise particularly and of food in all investigations 
similar to that undertaken by the writer. The ideal experiment would be to eliminate 
muscular exercise and food entirely, but this, of course, cannot be done (except at 
comparatively long intervals) and the body still be maintained in perfect health. It is to 
be regretted that a similar resting and fasting experiment was not made in the control 
period before leaving Ithaca for Scotland, and again immediately on arriving in 
Edinburgh. Any inherent diurnal rhythm which might be present would not, under 
these conditions, be masked by external influences, but I think the proof is 
sufficiently clear that there is no such rhythm. When the external conditions were as 
nearly as possible the same in the three localities mentioned, viz. the Orkneys, 
Winnipeg, and Ithaca, the curves are practically identical in character, and there is no 
indication of a Scottish rhythm being carried in the body to Winnipeg. In each of 
these curves (figs. 7 and 8) there is a slight morning rise and evening fall. The 
former comes in the period between sleep and full wakefulness, and the latter 
appears when daylight is replaced by artificial light, which again induces a feeling of 
drowsiness. 

During natural sleep the temperature of the body falls markedly, and much more so 
in the deeper sleep induced by narcotic drugs such as morphine, alcohol, etc. This is 
due in part to a diminution in muscular tonus which leads to a lessened heat- production, 
and in part to the fact that sleep is accompanied by a cutaneous vaso-dilatation 
allowing of an increase in heat-loss. Any condition which diminishes the activity of 
the heat-regulating mechanism, situated somewhere in the central nervous system, will 
tend to induce a state of poikilothermism, and then the temperature will fall if the 
temperature of the environment is lower (as it almost always is in temperate climates) 
than that of the body. Any tendency to somnolence, therefore, will be marked by 
some fall in the body temperature. 

If the curves of August 8 (Orkney) and August 29 (Winnipeg) be compared 
(figs. 7 and 8), it will be seen that in the former both the morning rise and the evening- 
fall are more distinct than in the latter. As explained previously, this is probably due 
to the fact that the subject was more quiescent on August 29. If the Scottish 
temperature rhythm had been carried in the body to Winnipeg, the morning rise and 
the evening fall should begin some six hours earlier ; the former, therefore, would not 
come into the Winnipeg curve at all, and the rectal temperature would be already high 
when the first observation was made at 6 a.m., while the ]atter would begin about 
12 (noon) or 1 p.m. instead of 6 or 7 p.m. As a matter of fact, a slight fall is 
noticeable in the Winnipeg curve at I p.m., but this is succeeded by a rise later which 
should not be if this corresponded with the Scottish evening fall. These two curves 
are remarkably alike in every respect. There are minor differences, but these are not 
greater than one might expect to find in any two similar curves taken in the same 
locality on consecutive days. 



daily routine and body temperature. 257 

VI. The Relative Values op Records from Rectum, Mouth, and Axilla, as 
indicating Changes in Body Temperature. 

As is correctly pointed out by Lindhard, the term "body temperature" is a 
misnomer. There is in reality no such thing. Most of the heat of the body is pro- 
duced in the muscles and in the organs of digestion, notably the liver, and from these 
it is conveyed by the blood to the other organs of the body. If by body temperature 
we mean the temperature of the warmest organ in the body, it will probably be found 
in the muscle or group of muscles which happens to be most active at that particular 
moment, and this will change from time to time according to circumstances. We may 
correctly speak of the rectal temperature, the mouth temperature, or the temperature 
of the axilla as the temperature of these localities at any particular time, but not of the 
body temperature. However, since the heat produced in the muscles, etc., is distributed 
so rapidly by the blood- stream that the temperatures of the deeper parts of the body 
away from the radiating surfaces do not differ at any one time probably by more than 
some fraction of a degree centigrade in health, the term body temperature may still be 
conveniently used as indicating the average temperature of the deeper and well- 
protected parts of the body. 

Clinically the temperature is usually taken in one of three situations — the rectum, 
mouth, or axilla, and of these it is recognised that the rectum is the best, since it is 
better protected against the rapid loss of heat than either the mouth or axilla, and 
consequently the readings are not so liable to be affected by external and accidental 
circumstances. The temperature of the rectum is always higher than that of the 
mouth and axilla and nearer the so-called body temperature. The mouth temperature 
is being used less and less by physicians and the axillary temperature scarcely at all, on 
account of the belief that these are unreliable as an indication of the changes in the 
temperature of the deeper parts, i.e. the body temperature. 

An examination of the charts (figs. 1 to 5) will show the relationship existing 
between the temperatures of these localities under different conditions. In the morning, 
before getting out of bed, the readings for the mouth and axilla are practically the same 
throughout the whole experiment, i.e. for the same day they are almost identical, both 
being distinctly below the rectum. After the subject arises the mouth temperature 
follows more or less closely the changes in the rectal temperature, the two curves 
running parallel to a considerable extent. The most constant deviation from the 
parallel is found between 9 a.m. and noon, when the mouth falls while the rectum 
rises, and this is easily explained. The 9 a.m. reading was taken shortly after 
breakfast, and the local effects of mastication and of warm food act on the mouth 
temperature alone. The rise thus produced quickly subsides, so that the next reading, 
at noon, is lower. 

The relationship between the curves of the axilla and rectum are much less constant. 
The temperature of the axilla almost invariably falls when the subject gets out of bed, 



258 DR SUTHERLAND SIMPSON ON 

while that of the mouth and of the rectum rises. The mouth and axilla curves separate 
at that point and do not meet again until the following morning. On the resting days, 
when the subject remained in bed, the figures for the mouth and axilla are practically 
identical at every observation, and the two curves run fairly parallel with that of the 
rectum. In the case of a patient confined to bed then, are the mouth and axillary 
temperatures as unreliable as they are held to be ? The temperature of the mouth 
particularly, and also of the axilla, is affected by the temperature of the surrounding air, 
rising and falling with it, and when this is fairly regular, as it usually is in a sleeping- 
room, the parallelism between the mouth and axillary curves and that of the rectum 
appears to be on the whole pretty constant. With the subject outside in the open air 
and exposed to atmospheric changes, the case, of course, is different. 

The above remarks, it must be remembered, are meant to apply only to the case of 
the person who was the subject of this experiment. There may be, and there probably 
are, great variations amongst different individuals in this relation. For example, in a 
sparely built person the walls of the buccal chamber will be thinner, and it will be more 
difficult to convert the axilla into a closed cavity than in a stout individual, so that the 
temperature of both localities will be more susceptible to changes in the surrounding 
air in the former than in the latter. 

With regard to the question of individual differences in the response of the mouth and 
skin temperatures to muscular activity I was interested to find, in reading over the report 
of Lindhard, # that his results were different from mine. He says : " With regard to the 
mouth temperature, its relation to the work of the muscles is quite inconstant. As a 
rule it does not rise during work. On working indoors, it is almost constant ; on work- 
ing in the open air it always falls, but in a single case 1 have found it so much raised 
while staying indoors soon after working in the open air, that the rise must without 
doubt be ascribed to the work. On the other hand, I have seen that even energetic 
exercise indoors was not able to prevent the mouth temperature, raised by the meal, 
from falling. . . . For the temperature of the skin pretty much the same holds good 
as for the mouth temperature ; as a rule I have not been able to notice any rise 
occasioned by muscular work." 

In order to test this matter further in the case of my own temperature I have lately 
made a few experiments, the results of which are given in tabular form below. These 
observations were made in September, when the outside temperature was fairly high, 
and again in November, when the weather was colder. The readings were taken first 
in the recumbent posture just before getting out of bedt in the morning, then in the 
sitting position half an hour later, after dressing but before breakfast, again after break- 
fast, and finally immediately after a walk of about three-quarters of a mile up a pretty 
steep hill and then two flights of stairs to a room in the laboratory. The same clothing 
was worn and the same muscular energy expended on each day in approximately the 

* LlNDHART), IOC. dt, p. 18. 

t The bed was outside on the verandah. 



DAILY ROUTINE AND BODY TEMPERATURE. 



259 



same time, so that the only variable factor was the air temperature. On the last two 
occasions breakfast was taken in the laboratory after the walk instead of before. 
The readings are given in degrees centigrade. 



TABLE IV. 
Effect of Muscular Action on the Rectal, Mouth, and Axillary Temperatures. 







Rectum. 


Mouth. 


Axilla. 


Air. 


Remarks. 


1911 














Sept. 7 


8 a.m. 


37-22 


3655 


36-52 


17 


In bed ; slight perspiration. 




8.30 „ 


37-32 


36-89 


36-31 


19 


After breakfast ; before dressing. 




9.30 „ 


37-42 


3676 


3630 


21 


Street car to laboratory. 


Sept. 8 


7.45 „ 


37-12 


36-56 


3645 


17 


In bed ; raining. 




8.30 „ 


37-41 


3652 


360 


20 


After dressing. 




9 » 


3738 


36-78 


3591 


20 


After breakfast. 




9.30 „ 


37-82 


3725 


3664 


15-20* 


After 17 minutes' walk. 


Sept. 9 


8 „ 


37-26 


3686 


36-70 


19 


In bed ; heavy rain. 




8.30 „ 


37-41 


37-04 


3670 


20 


After breakfast ; before dressing. 




9 


37-54 


3692 


36-72 


20 


After dressing. 




9.45 „ 


37-90 


37-24 


37-04 


19-20* 


Walk 16 minutes ; perspiring. 


Sept. 10 


9 


37-16 


3662 


36-60 


20 


In bed. 




9.30 „ 


37-18 


36-78 


36-70 


21 


After breakfast ; in bed. 




10 „ 


3751 


37 00 


36-54 


21 


After dressing. 


Sept. 11 


7 


3703 


36-62 


3662 


18 


In bed ; feeling cold. 




7.30 „ 


37-20 


3660 


36-18 


21 


After dressing ; before breakfast. 




8 


37-25 


3691 


36-38 


21 


After breakfast. 




8.45 „ 


37-72 


37-12 


36-78 


18-21* 


Walk 15 minutes ; perspiring freely. 


Sept. 12 


7.30 „ 


37-41 


37-00 


36-90 


20 


In bed ; slight lumbago. 




8 


37-60 


37-10 


36-64 


23 


After dressing ; before breakfast. 




8.30 „ 


37-75 


37-45 


36-72 


22 


After breakfast. 




9 


38-15 


37-65 


37-30 


18-22* 


Walk 16 minutes ; perspiring. 


Sept. 13 


7.30 „ 


37-00 


36-46 


36-46 


10 


In bed ; cold weather. 




8 


37 32 


36-48 


3566 


17 


After dressing ; before breakfast. 




8.30 „ 


37-36 


37-08 


36-28 


17 


After breakfast. 




9 


37-78 


37-12 


36-74 


9-19* 


Walk 15 minutes ; slight perspiration. 


Sept. 14 


6 


36-70 


36-03 


3600 


2 


Just after awoke ; feeling cold. 




7.30 „ 


36-78 


36-28 


36-12 


4 


In bed ; feeling cold. 




8 


36-92 


3631 


35-52 


17 


After dressing ; before breakfast. 




8.30 „ 


37-05 


36-50 


35-89 


17 


Alter breakfast. 




9.30 „ 


37-40 


3661 


3615 


5-16* 


Car to laboratory after business in town. 


Nov. 1 


7.15 „ 


36-94 


36-18 


36-20 


3 


In bed ; feeling warm. 




7.45 „ 


3725 


3638 


35-76 


20 


After dressing. 




t8.30 „ 


3757 


3621 


36-26 


2-22* 


Walk 15 minutes, talking on way, before breakfast. 




9 


37-42 


36-52 


36-26 


24 


After breakfast, in warm room. 


Nov. 3 


7.30 „ 


36-81 


36-48 


3636 


1 


In bed. 




8 


3712 


36-50 


35 82 


19 


After dressing. 




8.30 „ 


37-58 


36-71 


36-5 


1-21* 


Walk 15 minutes to laboratory before breakfast. 




9 


37-41 


36-68 


365 


21 


After cold breakfast. 



With regard, then, to the effect of muscular work on the mouth temperature, my 
results do not agree with those of Lindhard. On every day except one, the effect 
of a not very long walk was to raise the mouth temperature as well as the rectal, and 
usually the rise was very distinct. On all but two of these days the walk was begun 
very shortly after finishing a hot breakfast, when the mouth temperature was artificially 

* A thermometer was carried in the hand during the walk, and the lowest air temperature reached was noted ; 
this is indicated by the first figure, and the second gives the temperature of the room in the laboratory, where the 
observations were made immediately on entering it. 

t On this occasion I walked to the laboratory in company with a friend, with whom I was conversing all the way. 
On all the other days I breathed through the nose and kept the mouth closed. 



260 DE SUTHERLAND SIMPSON ON 

raised, and still, as a result of the exercise, it rose higher. During the walk the mouth 
was kept closed except on November 1, and that is the only occasion on which the 
temperature fell. The outside temperature has much to do with this, of course, and it 
is probable that in many of Lind hard's experiments it was below zero. He does not 
give the air temperatures, but when his observations were made indoors it is not likely 
that the atmosphere was as cold as that to which I was exposed in my walks on 
November 1 and 3, when the temperature was only one or two degrees above the 
freezing-point, and a cold wind blowing at the same time. Lindhard found in his own 
case and in other members of the crew that as a rule the mouth temperature did not 
rise during work. " On working indoors it was almost constant ; on working in the 
open air it always falls." The only explanation of the difference in the results is that 
it depends on individual peculiarities, or rather, on mouth peculiarities. The thinner 
the walls of the buccal chamber the greater and more rapid will be the dissipation of 
heat and the more difficult will it be to raise the temperature. The relative vascularity 
of the parts is probably also of importance. 

In my case the temperature of the axilla also rose during the walk, or at any rate 
did not fall. On November 1 and 3 it remained constant. 

VII. Summary. 

To determine whether the diurnal variation in body temperature is due to the 
combined effects of the various influences which are known to act upon it, such as 
muscular exercise, the ingestion of food, sleep, etc., or is present independently of 
these, the daily routine of the individual who is the subject of the experiment may 
be reversed artificially by causing him to work during the night and rest and sleep 
during the day, or it may be modified in another way, viz. by rapidly changing his 
longitude in a journe}' from west to east, or vice versa. If the temperature of the body 
is dependent on the influences mentioned, then a total reversal of the daily routine, 
or any modification of it, should produce a corresponding change in the diurnal 
temperature curve. 

On a voyage from Ithaca, in the western part of New York State, to Edinburgh, 
during six weeks' residence in Scotland, and again on the return journey from Edinburgh 
to Winnipeg, continuous three-hourly observations (and hourly from 6 a.m. to 9 a.m.) 
were made by the writer on his own temperature (rectum, mouth and axilla), except 
between midnight and 6 a.m., with a view to ascertain whether the temperature rhythm 
obeyed local (ship's) time or Ithaca time. To get the Ithaca rhythm as a control, 
observations were made in that city for one week before the journey began. If a 
temperature periodicity were fixed in the body independently of external conditions, 
then the curve should correspond to Ithaca time and not to local time as he travelled 
from west to east, or vice versa. For example, since Edinburgh local time is five hours 
in advance of Ithaca local time, the morning rise which began at 7 a.m. in Ithaca 



DAILY ROUTINE AND BODY TEMPERATURE. 261 

should have been delayed till 12 (noon) in Edinburgh, if the Ithaca rhythm were fixed 
in the body of the subject. This was found not to be the case. In the eastward 
voyage, which occupied eight days, there was an immediate adjustment of the 
temperature rhythm to the changed routine day by day, and the same was found to 
occur on the westward journey from Edinburgh to Winnipeg, between which stations 
there exists a difference of over six hours in time. 

It was also found that the temperature curve can be almost obliterated by remaining 
in bed and enforcing muscular rest during the whole twenty-four hour period, at the 
same time abstaining from food. This was practised in Scotland, in Winnipeg, and 
in Ithaca. Any temperature rhythm inherent in the body might be expected to show 
itself under these conditions, and, if persistent, to be carried from one locality to the 
others. This was found not to be the case. The curves obtained at the above- 
mentioned stations, with time differences as already indicated, resemble each other 
very closely, and differ only in minor details. 

The results of the present investigation, therefore, give strong support to the 
conclusions of Gibson, Lindhard, and Simpson and Galbraith, viz. that the diurnal 
variation of body temperature, in man as in other animals, is determined by the 
conditions imposed on the body, such as rest and activity, and is not an expression 
of any inherent periodicity established in the body. 



( 263 ) 



XIII. — On the Carboniferous Flora of Berwickshire. Part L Stenomyelon Tuedianum 
Kidston. By R. Kidston, LL.D., F.R.S., and D. T. Gwynne-Vaughan, M.A., 
Professor of Botany, Queen's University, Belfast. (Plates I.-IV.) 

(MS. received January 8, 1912. Read January 8, 1912. Issued separately April 13, 1912.) 

Introduction. By R. Kidston. 

My first knowledge of Stenomyelon Tuedianum was derived from a microscopical 
preparation in the collection of the late Mr C. W. Peach, A.L.S., which, through the 
kindness of his son, Dr B. N. Peach, F.R.S., subsequently came into my possession.* 
The specimen was labelled as having been found "near Berwick." 

During the investigation of the fossil flora of Berwickshire, in company with Mr 
A. Macconochie of the Scottish Geological Survey, a special effort was made to secure 
additional specimens of this plant, as the original material, as far as known to me, 
was quite inadequate for a satisfactory description of the species. It was ascertained 
that the specimen in Mr Peach's collection had been received from the late Mr Adam 
Matheson, Jedburgh, a geologist who took much interest in the fossils of his neighbour- 
hood, and whom I believed to be the author of an anonymous pamphlet — apparently 
a reprint from a local newspaper — describing some fossil stems found at Norham 
Bridge. A perusal of this paper led me to infer that his specimen of the fossil, here 
described as Stenomyelon Tuedianum, had also come from the same locality, an opinion 
which subsequent events showed to be correct. 

The matrix containing Mr Matheson's fossil was an impure fine clay, apparently 
with a fair proportion of iron, and one showing features which were possible of recognition 
in the field ; but though a careful search for a similar bed was made in the neighbourhood 
of Norham Bridge, no trace of such could be found in situ. Subsequently, in 1901, 
we discovered some small blocks of the desired rock lying on the side of the road 
near the north end of Norham Bridge. It was ascertained that the material came 
from a cutting made in the road while putting in a drain some time before ; the 
surface of the road in the neighbourhood of the drain was therefore carefully examined, 
and in a small block which had been used for refilling the cutting the specimen was 
discovered which has enabled us to give a detailed description of the species. 

We are also indebted to the original material for additional points of interest. 

Subsequently, in 1903, Dr D. H. Scott, F.R.S., informed me that some sections of 
Stenomyelon had been presented in 1859 to the Botanical Museum, Royal Botanic 
Gardens, Edinburgh, by Mr Adam Matheson ; and from a copy of a letter referring to 
these specimens which has been kindly forwarded to me by Dr Scott, who had received 

* Slide No. 2105. 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 13). 40 



264 DR R. KIDSTON AND PEOFESSOR D. T. GWYNNE-VAUGHAN ON 

it from Professor I. Bayley Balfour, F.R.S., it would appear that Mr Matheson only 
found one specimen of Stenomyelon, and that this had been got at Norham Bridge, 
which the study of the anonymous paper already mentioned had led me to infer. 

In addition to the parts of the original specimen received from Mr Matheson by 
Professor Balfour and Mr C. W. Peach, in 1900 I received a few fragments attached 
to glass slips from Miss Matheson, daughter of the original discoverer of Stenomyelon 
Tuedianum. Some additional specimens, also mounted for microscopical examination, 
were given me by Mr James Veitch, Jedburgh, to both of whom I take this opportunity 
of recording my sincere thanks. 

Description of Stenomyelon Tuedianum, Kidston. By R Kidston and 

D. T. Gwynne-Vaughan. 

Owing to the imperfect preservation of the softer tissues of the specimen it is 
not possible to determine the true outline of the stem, for in all cases the thin-walled 
tissues of the cortex are completely disintegrated and destroyed. As a result, the 
stem has become flattened, and those tissues of the cortex that remain form wing-like 
projections on two opposite sides of the stele. One of these extensions is seen in its 
entirety, but of the other only a part is preserved. In some of the sections it is quite 
narrow, having a regular and very wing-like appearance (figs. 1-2), while in others 
it is much wider, and is irregularly crushed in at its margin (fig. 3). The xylem of 
the stele has not suffered any damage, and is almost circular in section. We are 
of the opinion that the present appearance of the fossil does not imply that the stem 
of the living plant actually possessed wings, but it will be shown that the leaves were 
arranged in three rows, and it is therefore quite possible that the stem possessed three 
more or less prominent decurrent ridges. 

Description of the Stem. 

The xylem mass of the stem contains both primary and secondary xylem, and as a 
whole measures 8-9 mm. in diameter. The primary xylem forms a bluntly triangular 
or three-lobed central mass, which measures 3-4 mm. from the extremity of one lobe to 
that of another (figs. 6-10). It consists of tracheae alone, except for a few occasional cells 
of parenchyma near its periphery. There are also three bands of parenchymatous tissue 
in the primary xylem which extend from the middle of the sides of the triangle so as to 
separate the three lobes more or less completely from one another. These parenchymatous 
stripes, although always quite narrow, vary somewhat in breadth, and they also form a 
few small irregular bays projecting into the tracheal masses of the lobes. Sometimes, 
but rarely, the three stripes are continuous to the centre, and then the lobes of the xylem 
are completely separate from one another. More often one or the other of them is 
interrupted by a wider or narrower bridge of tracheae, so that two of the lobes are con- 



THE CARBONIFEROUS FLORA OF BERWICKSHIRE. 265 

nected. In spite of these occasional fusions, the three lobes of the primary xylem very 
clearly retain their individuality throughout all the sections, as may be seen by comparing 
the above-mentioned figures, — so much so indeed that they may be regarded as practically 
independent of one another. 

These three lobes are obviously in relation to the leaf-traces, one of which, as will be 
shown later, departs from the extremity of each lobe in turn. It is clear, therefore, that 
the divergence must have been ^. The sequence in which fusion occurs between any 
two of the lobes is possibly also related to the order of the leaf-trace departure, but our 
series of sections was not sufficiently extensive to settle this point. 

The elements of the primary xylem are fairly large, attaining a diameter of 160/*. 
In longitudinal section they are seen to be typical elongated tracheae with finely pointed 
ends, and beautiful porose pitting on all their vertical walls (fig. 12). Towards the ex- 
tremities of the lobes the tracheae become somewhat smaller, and their pitting becomes 
typical scalariform (fig. 13). There appear to be no protoxylem strands proper to the 
stem itself, but in the neighbourhood of the departing leaf-traces a pair of definite exarch 
protoxylem strands are usually to be seen (fig. 11, prx.). These are evidently decurrent 
from the leaf- trace, and lower down in the stem they fuse to a single strand (fig. 10, prx.). 

The secondary xylem evidently appears first opposite the side bays of the triangle of 
the primary xylem, and when the bays are filled up it crosses over in front of the angles, 
so that the outline of the whole xylem mass, which was at first triangular, eventually 
becomes circular. The earlier elements of the secondary xylem are narrower than those 
formed later (fig. 11 and figs. 6-10), and its tracheae are usually separated from those of 
the primary xylem by a few cells of parenchyma. Often, however, primary and secondary 
tracheae are found to be in direct contact. The tracheae of the secondary xylem are 
arranged in very regular radial rows, which are separated by the medullary rays into 
wedges of from 2-5 elements in width (fig. 11). The medullary rays widen out con- 
siderably towards the periphery, and in this region secondary medullary rays also make 
their appearance. The rays are all very long vertically, and are from 1-6 cells broad 
(fig. 14). A. few specially wide medullary rays occur on the inside of the departing leaf- 
trace (fig. 10, m.r.). In radial section the cells of the rays have the ordinary brick-like 
appearance, with the long axis radial (fig. 15). All the tracheae of the secondary wood 
exhibit porose pitting, but the pits are confined to the radial walls ; the tangential walls 
are quite smooth, being entirely without pits (fig. 14). 

No trace was found of any tissue resembling phloem, neither around the stele of the 
stem nor in relation to the leaf-traces in the cortex. However, patches of a very curious 
tissue have been preserved in the immediate neighbourhood of the stele, which are clearly 
parts of a zone that in life completely and closely surrounded it. The inner portion of 
this tissue consists of small cells with very dark contents. They are rectangular in form, 
and are arranged in very regular radial series. In longitudinal section they have 
exactly the same appearance as in transverse, only their long axes are vertical instead 
of being tangential (fig. 16, i.z.). There can be no doubt that the cells arose from the 



266 DR R. KIDSTON AND PROFESSOR D. T. GWYNNE-VAUGHAN ON 

activity of a cambium of some kind, and the general appearance and structure of the 
tissue indicates it is best regarded provisionally as a deep-seated and probably thick- 
walled periderm. A certain amount of phloem and other soft tissues may have existed 
between this zone and the xylem, but if so they have entirely disappeared. In any 
case such tissues could not have been very extensive, for, after making every allowance 
for crushing and imperfect preservation, the dark-celled zone fits very closely into the 
xylem, — so much so indeed at some points as almost to suggest that it might have arisen 
from the same cambium. 

Towards the outside of this dark-celled zone its elements become larger and 
irregularly polygonal in outline, the radial arrangement being in consequence lost 
(fig. 16, o.z.). In addition, large groups of similar cells form rounded bosses, projecting 
at irregular intervals beyond the general level of the outer surface of the zone (fig. 1 6, 
b.). The same type of cell also forms a number of scattered sphserical masses in the 
cortex of the stem (fig. 17). The central cells of these masses, and also those of the 
bosses, are smaller than the others, and form a sort of nucleus from which the larger cells 
radiate more or less distinctly, appearing as though, during their development, they had 
been pulled out by the growth of the surrounding tissue. Their cell walls are very im- 
perfectly preserved, but it is very probable that they were of a sclerotic nature. It is 
clear that the masses they form were during life quite hard and resistent, for whenever 
they have been pressed against a mass of xylem, it is the latter, and not the former, that 
has given way and become crushed. The " sclerotic nests," as they may be called, are 
scattered without order throughout the inner cortex of the stem, and do not seem to 
bear any relation to the leaf-traces. In structure and appearance they are quite similar 
to the sclerotic nests described by Dr Scott in the pith of Calamopitys Beinertianum* 

The outer cortex of the stem is of the " Sparganum " type (fig. 18). That is to say, 
it contains a number of rather short radiating bands of fibrous sclerenchyma, running 
vertically almost without anastomosis (fig. 19, from the petiole). The tissue between 
these sclerotic bands is a rather firm-walled parenchyma. Some of its cells have dark- 
coloured contents, and may perhaps be regarded as resin sacs. 



The Departure of the Leaf-trace. 

The leaf-traces depart from the ends of the lobes of the primary xylem in a perfectly 
protostelic manner. Our series enabled us to follow one such departure from start to 
finish. First the extremity of one of the xylem lobes is seen to become more and more 
prominent (figs. 6-7, l.t. 2 ). Then the protrusion is gradually nipped off as a fairly large 
roundish leaf-trace (figs. 8-9, l.t. 2 ). As the trace passes out it carries with it some of 
the secondary xylem on its abaxial side, and at the same time it also obtains a certain 
amount of secondary xylem of its own on the adaxial side (figs. 9-10). The depar- 

* Scott, " On the Primary Structure of certain Palajozoic Stems with the Dadoxylon type of Wood," Trans. Roy. 
She. Edim., vol. xl. p. 342, pi. i. figs. 3-4, 1902. 



THE CARBONIFEROUS FLORA OF BERWICKSHIRE. 267 

ture of the trace leaves no gap in the secondary xylem, which closes in immediately 
behind it. 

The leaf-trace continues to possess a certain amount of secondary xylem for some 
time after it has become free from the xylem of the stem, although that on the adaxial 
side is very scanty and soon disappears. It was possible also to recognise in our sections 
the leaf-trace that immediately preceded the one just described (figs. 6-7, l.t., 1 and 18). 
In this trace the primary xylem had just divided into two approximately equal portions, 
which are at first still surrounded by a certain amount of secondary xylem common to 
both. Even after they have become completely separate, the secondary wood on the 
abaxial side is retained for a considerable time (fig. 7, l.t. 1 ). 

There is no doubt that all the traces occurring in the space formerly occupied by the 
cortex were derived by the continued subdivision of primary leaf-traces such as those 
just described. Several leaf-traces, both large and small, were observed in actual division 
(fig. 21). It is therefore certain that all the vascular strands given off from the stele 
are primary leaf-traces, which in their passage through the cortex divide repeatedly to 
supply the petioles. There is also clearly an entire absence of any meristeles proper to 
the stem such as those occurring in Sutcliffia insignis* 

The xylem of the leaf-traces consists mainly of porose tracheae without any xylem 
parenchyma. The tracheae at the periphery are markedly smaller than those at the 
centre, and are scalariformly pitted (figs. 20,21, 22, and 23, from the petiole). The 
scalariform tracheae are present in greatest number towards the side of the trace on which 
the protoxylem is situated. The traces usually show a single very distinct and definite 
protoxylem, except just below a point of division of the trace where there are two (figs. 
21-22). In the immediate neighbourhood of the stele the protoxylems appear to be 
exarch, but at some distance out they are distinctly immersed, and are represented by 
small cavities situated some little distance within the periphery of the xylem (figs. 20, 
21, and 22). Owing to the shifting and displacement of the leaf- traces due to collapse 
of the cortex, it is impossible to determine their proper orientation with regard to the 
periphery of the stem. Most probably, however, the protoxylems pointed to the outside. 
No indication of phloem could be discovered in relation to any of the leaf-traces. 



Description of the Leaf. 

In the sections of the block which contained the stem there also occurred several other 
fragments of vegetable tissue. Most of these were portions of leaf lamina, and one of 
these was found to be in organic connection with an axis which is no doubt the petiole, 
rachis, or midrib of Stenomyelon (fig. 1, l., and text fig. 1). As seen in longitudinal 
section, the connection between the lamina and the rachis of the petiole extended over 
a considerable distance, which suggests that the leaf was a simple and not a divided one, 

* Scott, " On Sutcliffia insignis, a new type of Medullosese from the Lower Coal Measures," Trans. Linn. Soc, vol. vii. 
p. 49, 1906. 



268 



DR R. KIDSTON AND PROFESSOR D. T. GWYNNE-VAUGHAN ON 



and from the portion of the leaf contained in the block it is evident that it was con- 
tinued below into a round or oval petiole, which bore the curious emergences described 
later on (text fig. 1 ). 

The leaf is very large compared with the stem that occurs alongside of it, and it 
is possible that it belonged to another and perhaps a larger stem (fig. 1, l.). In the 
section only about one-half of the total circumference of the midrib or rachis was 
present, and it contained only four vascular strands. These are of exactly the same 
type as the leaf-traces in the cortex of the stem, but they are on a distinctly larger 
scale. (Compare fig. 25 with figs 20, 21, and 22, which are all of the same magni- 
fication.) The small peripheral elements are more numerous in the strand of the leaf, 
and in longitudinal section there is a greater proportion of scarlariform to porose 




Fig. 1. — Stenomyelon Tuedianum. Free petiole showing vascular strands, " Spai'ganum " cortex, and emergences, x 8i. 

(Slide No. 2095.) 

elements. Some of the scalariform elements are almost as wide as the porose, but 
transitional types of pitting are to be observed (fig. 23). The protoxylems are slightly 
immersed, and in one case two were present, indicating an approaching division 
(fig. 25). In the three strands that were sufficiently well preserved to show the 
protoxylems they were on the side of the xylem nearest the periphery of the midrib 
or rachis. The outer cortex is of the " Sparganum" type (figs. 28 and 19). It attains 
a greater development than that seen in the stem, — the sclerotic strands being much 
more numerous and more irregular in form. The dark-coloured resin sacs that occur 
between the fibrous strands in the stem are present in greater number in the midrib, 
and they also occur in its more central parenchyma. The "Sparganum" cortex does 
not reach up to the extreme periphery of the petiole, but there is a narrow zone of 
homogeneous parenchyma lying to the outside of it (figs. 31 and 4, par.). At certain 
points this tissue is prolonged into a number of curious emergences (figs. 30, 31, and 
4). They vary greatly in their form and length, some being quite short and blunt 



THE CARBONIFEROUS FLORA OF BERWICKSHIRE. 269 

(fig. 31), while others are narrow, greatly elongated, and somewhat swollen at the end 
(figs. 30 and 4). They do not seem to be produced by any crushing or folding of 
the outer tissues. It is possible, however, that they may represent transverse sections 
of longitudinal ridges. The epidermis of the midrib is well preserved as a definite 
layer of rather small cells. 

Judging from a fragment present in the original block found by Mr Matheson, 
the lamina of the leaf must have been of considerable thickness (fig. 24). The 
epidermis is quite distinct, but no stomata could be distinguished in it. There were 
indications of a hypodermal zone of sclerotic strands on both sides of the leaf, and 
resin sacs were very plentiful in the mesophyll. In the fragment in question the 
vascular bundles lie in single row, but in others less well preserved they appear to be 
irregularly scattered. 

Remarks on other Stems of Stenomyelon. 

A few other stems were present in Mr Matheson's material which in all essential 
points resemble the specimen already described, but they exhibit a few features of special 
interest, to which it is necessary to refer. 

In one of these stems the primary xylem at first sight gives the impression of being 
parenchymatous (figs. 26-27). We were able to satisfy ourselves, however, that this is 
not the case. The apparent parenchyma cells are really tracheae that have undergone 
a certain amount of decay. In this section also the three lobes of primary xylem 
appear to be quite separate from one another, but this, again, is probably due to 
decay (fig. 26). 

In another stem (fig. 29) the secondary thickening has only just begun. It is 
present in greatest quantity opposite the sides of the triangle of primary xylem, and has 
not yet crossed over in front of all the lobes. 

Again, an unusually small stem was found in which the primary xylem only measures 
2*5 mm. at its greatest width (fig. 5), while the secondary xylem is much more ex- 
tensive than it is in the stem shown in fig. 29, which is photographed at the same 
magnification. This indicates that, apart from secondary thickening, the primary axes 
of Stenomyelon were of different sizes, which might perhaps suggest that the plant was 
branched. 

Conclusion. 

The stem of Stenomyelon possesses so many features peculiar to itself that in the 
present state of our knowledge it is unsafe to enter into any speculation as to its relation- 
ship to other members of the Cycadofilices. It is perhaps best to let it remain among 
that nebulous group in which it has been already provisionally placed by Dr Scott.* 
At the same time it should be noted that the absence of independent meristeles in the 

* Studies in Fossil Botany, 2nd ed., part ii. p. 498, 1909. 



270 DR R. KIDSTON AND PROFESSOR D. T. GWYNNE-VAUGHAN ON 

cortex of the stem separates it widely from Sutcliffia insignis, with which one might be 
tempted to compare it. 

Locality. — Eoad Cutting at north end of Norham Bridge, Berwickshire. 

Horizon. — Calciferous Sandstone Series. 

Diagnosis. 

Stenomyelon Tuedianum, Kidston, gen. et spec. nov. 

Stem monostelic, primary xylem without xylem parenchyma, divided more or less 
distinctly into three lobes by as many radiating and interrupted bands of parenchyma. 
Primary tracheas porose on all walls. The protoxylems of the leaf-trace decurrent as 
exarch strands on the extremities of the lobes. Secondary thickening occurs. Secondary 
tracheae, with porose pits on radial walls only. Medullary rays numerous. Stele closely 
invested by a zone of sclerotic periderm. Leaf-traces depart successively from the 
extremities of the lobes and repeatedly divide in the cortex. Leaf-trace protoxylems 
become immersed. Outer cortex of the " Sparganum " type. 



DESCRIPTION OF PLATES I.-IV. 

Plate I. 

Stenomyelon Tuedianum, Kidston. 



Fig. 1. General view of the specimen, showing a stem, st., and a portion of a leaf, l. x2|. (Slide 2093.) 
Fig. 2. Transverse section of the stem, showing the cortex flattened into a thin wing. x 2£. 

(Slide 2089.) 

Fig. 3. Transverse section of the stem, showing a thicker and indented cortex. x 2|. (Slide 2097.) 
Fig. 4. Transverse section of outer cortex of midrib of leaf, showing two long emergences; par., 

parenchymatous cortex. x 13. (Slide 2094.) 

Fig. 5. Transverse section of the small stem in Mr Matheson's original block. x 9. (Slide 2107.) 
Fig. 6. Transverse section of stem, showing the beginning of the departure of a leaf-trace, It. 2 , x 9. 

(Slide 2088.) 

Fig. 7. Do., It. 1 , the trace of the proceeding leaf already divided into two. x 9. (Slide 2092.) 

Plate II. 

Fig. 8. Transverse section of stem, showing leaf-trace just free from primary xylem. x 9. 

(Slide 2095.) 

Fig. 9. Do., leaf-trace passing through secondary xylem. x 9. (Slide 2096.) 

Fig. 10. Transverse section of stem, showing leaf-trace free from secondary xylem; m.r., specially large 

medullary rays below the leaf-trace ; prx., single protoxylem at summit of lobe, x 9. (Slide 2098.) 

Fig. 11. Transverse section of a lobe of the primary xylem with two decurrent leaf-trace protoxylems, 

prx. x 35. (Slide 2091.) 

Fig. 12. Longitudinal section of central tracheae of primary xylem. x 110. (Slide 2101.) 

Fig. 13. Longitudinal section of periphery of primary xylem, showing small scalariform elements. 

xllO. (Slide 2099.) 



THE CARBONIFEROUS FLORA OF BERWICKSHIRE. 271 



Plate III. 

Fig. 14. Tangential section of secondary xylem, showing medullary rays and unpitted surface of the 
tangential tracheal walls. x 110. (Slide 2099.) 

Fig. 15. Radial section of secondary xylem, showing porose tracheae and medullary rays. x 110. 
(Slide 2100.) 

Fig. 16. Radial section at the periphery of the stele, showing its periderm; i.z., inner zone; o.z , outer 
zone of the same; b., the hosses; xy. 2 , secondary xylem. x 30. (Slide 2099.) 

Fig. 17. Longitudinal section of the inner cortex, showing the sclerotic nests. x 30. (Slide 2101.) 

Fig. 18. Transverse section of the cortex of the stem, showing the "Sparganum" zone; l.t., 1 same leaf- 
trace as in fig. 7. x 10. (Slide 2088.) 

Fig. 19. Tangential section through the "Sparganum" zone of the midrib. x 17. (Slide 2102.) 

Fig. 20. Transverse section of a leaf-trace in the cortex of the stem, showing a single protoxylem, 
prx. x 36. (Slide 2096.) 

Fig. 21. Do. The trace in division. x 36. (Slide 2097.) 

Fig. 22. Do. Below the point of division there are two protoxylems. x 36. (Slide 2090.) 

Fig. 23. Longitudinal section of the xylem of a leaf-trace in the midrib, showing' porose, scalariform, 
and transitional pitting. x 110. (Slide 2100.) 

Plate IV. 

Fig. 24. Transverse section of the lamina of a leaf. x 13. (Slide 678.) 

Fig. 25. Transverse section of a vascular bundle in the midrib near a point of division ; prx., protoxylem. 
x 36. (Slide 2094.) 

Fig. 26. Transverse section of a stem in Mr Matheson's block, with false appearance of parenchyma 
in the primary xylem. x 9. (Slide 678a.) 

Fig. 27. Portion of the same. x 30. (Slide 678a.) 

Fig. 28. Transverse section of "Sparganum" cortex of the midrib, cf, fig. 19; scl., sclerotic cortex; 
par., parenchymatous cortex, x 15. (Slide 2094.) 

Fig. 29. Transverse section of a stem in Mr Matheson's block, showing the beginning of secondary 
thickening, x 9. (Slide 2105.) 

Fig. 30. Transverse section of outer cortex of midrib, showing a long emergence. x 13. (Slide 2093.) 

Fig. 31. Transverse section of outer cortex of midrib, showing several short emergences ; scl., sclerotic 
cortex ; par., parenchymatous cortex. x 13. (Slide 2095.) 

(All the figured specimens are in the collection of Dr R. Kidston.) 



TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 13). 41 



Roy. Soc. Edin r 



Vol. XLVIII. 



KlDSTON AND GwYNNE-VaUGHAN. — PLATE I. 




i, Photo. 



Stenomyelon Tuedianum, Kidston. 



M'Farlane & Erskine, Lith., Edin , 



an s Roy. Soc. Edin r - 



KlDSTON AND GwYNNE-VaUGHAN.-PlATE II. 



Vol. XLVIII. 




p^^l 



I '^ M ' M »' * ?*«« »■** * ■» , x^t I^VVA^ t p 

^ V*. 

I - 



'''WHWW 



mtffl*pff ff ffr k '* !# " l, t *T?r°** ' '*~' 







- 



,*^lpli^!i^cP^,i 



!*■■■■ 



12 



. . .f . . . . 



'in, Photo 





'^^w 1 ' ^^B mi"*" 



prx 



prx 



\ n\ 1 "^W- . .... ..,. „„., .„ ! I TT„T; 



>'. <tHWMM%fM».. . ••■ 'HIMbi;,.: 



« j|^M>;>- Mi 



■*V« • • »».'i 



13 



Stenomtelon Tuedianum, Kidston. 



M'Farlane & Erskine, Lith., Edin.. 



itoy. Soc. Edin r 



Vol. XLVIII. 



KlDSTON AND GwYNNE-VaUGHAN.- PLATE III. 




Stenomyelon Tuedianum, Kidston 



M'Farlane & Erskine, lath., Edin , 



ms. loy. Soc. Edin r - 



Vol. XLVIII. 



KlDSTON AND GwYNNE-VaUGHAN.— PLATE IV. 




"i>hoto. 



Stenomyelon Tuedianum, Kidston. 



M'Farlane & Erskine, Lith., Edin., 



( 273 ) 



XIV. — The Cephalopoda of the Scottish National Antarctic Expedition. 
By William Evans Hoyle, M.A., D.Sc. 

(MS. received January 8, 1912. Read February 19, 1912. Issued separately May 28, 1912.) 

The Cephalopoda collected by the Scotia may, with a few trifling exceptions, be 
separated geographically into three divisions, coming respectively from South Africa, 
South America, and the Antarctic. 

A. South Africa : 

Ewprymna sp. 
Sepiolid (undetermined). 
Loligo reynaudi. 
Sepia australis. 
Hemisepius typicus. 

B. South America : 

Polypus brucei, n. sp. 
Polypus tehuelchus. 
Desmoteuthis sp. 

C. Antarctic Regions : 

Stauroteuthis sp. 
Moschites charcoti. 
Onyehoteuthis ingens. 
Bathyteuthis abyssicola. 
Galiteuthis suhmi. 

In addition were collected : — 

Between the Cape and Tristan da Cunha : 
Histioteuthis sp. 

Equatorial Atlantic : 

Tremoctopus quoyanus. 

A considerable number of the horny mandibles of Cephalopods were obtained from 
the stomachs of various mammals and birds, but the small amount of authentically 
TRANS. ROY. SOC. EDIN., VOL. XLVIIL, PART II. (NO. 14). 42 



274 DR WILLIAM EVANS HOYLE ON THE 

named material available does not justify an attempt to identify them. The animals 
referred to and the localities were : — 

Ross' Seal.— Station 165, 6th February 1903. 

Weddell's Seal.— Station 326, Jessie Bay, South Orkneys, May 1903 ; Station 325, 
South Orkneys, 21st September 1903. 

Albatross.— Station 437, 3rd April 1904. 

Sooty Albatross.— Station 376, lat. 64° 38' S., long. 35° 13' W. 23rd February 
1904. 

Emperor Penguin.— Station 248, lat. 69° 46' S., long. 20° 58' W. 21st February 
1903. 



SYSTEMATIC LIST. 

ClRROTEUTHID^. 

Stauroteuthis sp. 

Locality.— Station 295, Weddell Sea. Lat. 66° 40' S., long. 40° 35' W. 10th March 
1903. 2425 fathoms. One specimen [H 956].* 

This is probably either S. meangensis or S. hippocrepium, but in the mutilated 
condition of the body and the absence of the internal cartilage it is impossible to 
speak with certainty. It is just possible that it might be one of the species of 
Cirroteuthis, but this is less likely. 

A number of fragments and a few fairly complete examples of Crustacea were 
found in the gizzard of this specimen, and an account of them has been published by 
Dr Thomas ScoTT.t The most remarkable appears to be Poiitostratiotes abyssicola, 
O. S. Brady, which seems never to have been met with since the unique type was 
obtained by the Challenger in mud from 2200 fathoms in lat. 37° 29' S., long. 
27° 31' W. This is of interest as furnishing corroborative evidence of the deep-sea 
habits of the Cirroteuthidse. By a clerical error Dr Scott gives the date of capture 
as 1908 instead of 1903. 

A water-colour drawing of this specimen, made on the Expedition, shows that the 
coloration very closely resembles that of Stauroteuthis hippocrepium, as depicted 
in the Albatross Report ;j the colour of the body is, however, more deeply purple. 
As compared with Cirroteuthis umbellata, Fischer, § the arms are dull red instead 
of deep purple (though this may be owing to the oral aspect of the arms being 
depicted in one case and the aboral in the other), and the body is purple instead 
of pale reddish. 

* The numbers in square brackets refer to my own register of specimens examined. 
t Ann. and Mag. Nat. Hist. (8), vol. v. pp. 51-54, pis. ii., Hi., .Ian. 1910. 
X Hoyle, Bull. Mus. Comp. Zool., vol. xliii., No. 1, pi. i. fig. 1, pi. ii. fig. 1, 1904. 
§ Joubin, "Ce|ihalopodes <le la ' Princesse Alice,'" pi. i., 1900 [1901]. 



CEPHALOPODA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 275 




276 



DR WILLIAM EVANS HOYLE ON THE 



Tremoctopodid^e. 

Tremoctopus quoyanus, d'Orbigny, 1835. 

Locality.— Tow-net, Station 59, Equatorial -Atlantic. Lat. 2° 30' S., long. 32° 42' W. 
12th December 1902. Surface. One specimen, $ [H 1366]. 
Previous Records. — Atlantic and Pacific Oceans. 

Polypodid^. 

Polypus brucei, n. sp. 

Locality. — Station 346, Burdwood Bank, off Tierra del Fuego. 1st December 
1903. One specimen, $ [H 924]. 

The Body is a flattened ovoid, with a very shallow groove along the middle 





Fig. 2. — The hectocotylised arm of Polypus brucei. a, oral aspect of 
the. extremity. Natural size. 



line ventrally. The mantle opening extends fully half way round the circumference 
of the body, terminating immediately below and behind the eyes. The siphon is 
short and broad, and extends less than half way from the margin of the mantle to 
the edge of the umbrella. 

The Head is somewhat narrower than the body, and the eyes are but slightly 
prominent. 

The Arms are somewhat unequal, and about four times as long as the body ; 
their order of length is 1, 2 = 3, 4. The umbrella is well marked and its arrange- 
ment very characteristic. On the dorsal aspect of each arm it is attached as far as 
a point about one-third up the arm, whilst on the ventral aspect its attachment 
can be followed to about within 1 cm. of the extreme tip of the arm. The suckers 



CEPHALOPODA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 277 

(fig. 1) on all the eight arms are enlarged for the second quarter of the arm ; after 
about the first twelve suckers they enlarge very rapidly for about six suckers, and 
then gradually diminish. The third arm on the right side is hectocotylised (fig. 2), 
and is considerably shorter than its fellow on the opposite side. The seminal groove 
is well marked, but is neither very broad nor very deep ; the modified extremity 




Fig. 3. — Radula of Polypus brucei [H 924]. x 25. 

is unusually long and narrow, and, instead of the usual transverse ridges, shows a 
double row of small papillae along its bottom. 

TJie Surface shows a considerable amount of wrinkling, due apparently to the 
action of preservative fluids, but was most probably smooth when the animal was 
alive. There is no trace of any warts or tubercles. 

The Colour is dull purplish above, changing gradually into a pinkish stone 
colour below. 

The Radula is shown in fig. 3. 



Dimensions in Millimetres. 



End of body to mantle margin . 








58 


End of body to eye 








75 


Breadth of body 








60 


Breadth of head 








50 


Eye to edge of umbrella 








60 


Length of hectocotylus . 








17 


Breadth of hectocotylus 








3 


Diameter of largest sucker on arm 








15 






Right. 


Left. 


Length of first aim 




270 


275 


Length of second arm 




185* 


250* 


Length of third arm 




200 


260 


Length of fourth arm 




255 


255 




* Mutilated. 









This species is evidently nearly related to P. megalocyathus (Gould) from the 
same geographical region. It differs, however, in the absence of the extremely 



278 



DR WILLIAM EVANS HOYLE ON THE 



marked constriction between the head and umbrella, as well as of the membrane 
along the sides of the body, and in the fact that the enlarged suckers are found 
in all the arms. It is impossible to ascertain whether this last peculiarity occurs 
in Gould's species, but his comparison with P. fontania?ius, in which only the 
lateral arms have enlarged suckers, would lead one to suppose that such was the 
case in his species also. 

I have much pleasure in dedicating this species to my friend Dr W. S. Bruce, 
the leader of the expedition. 

Polypus tehuelchus, d'Orbigny, 1835? 

Locality.— Station 118, Falkland Islands. Lat. 51° 49' S., long. 57° 51' W. 
Shore collection. 6th February 1904. One specimen, $ [H 1696]. 





Fig. 4. — Hectocotylised arm of Polypus tehuelchus. 
a, oral aspect of tlie extremity. Natural size. 



Port Stanley, Falkland Islands. February 1904. One specimen, £ [H 926]. 

Previous Records. — -East coast of Patagonia, 40° S. ; Strait of Magellan ; Punta 
Arenas ; Nicaragua ; St Thomas, Danish West Indies. 

The skin of the upper part of the body, and especially of the head, is very 
much wrinkled, but this is probably due to the action of reagents, as no traces of 
definite papillae can be found. The animal was most likely smooth in the natural 
state. The hectocotylised arm of the male (fig. 4) has a very well-developed 
seminal groove, especially at the proximal end, where the membrane forming it 
stands out very distinctly from the surface of the arm. The tip is comparatively 
short and broad, measuring 6x3 mm., and of quite normal form ; the terminal 
groove is small and narrow ; its margins are deeply folded (perhaps owing to reagents), 
and there are no transverse ridges across its bottom. The radula is shown in fig. 5. 

I believe this specimen to be correctly identified, but there is some little doubt 



CEPHALOPODA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 279 

owing to its colour being neither so dark above nor so pale below as is indicated 
in d'Orbigny's description and figure. 




Fig. 5.— Radula of Polypus tehuelchus, 6 [H 1696]. x 52. 



Moschites charcoti (Joubin), 1905. 



Locality — Station 325, Scotia Bay, South Orkneys. 7th August 1903. 10 
fathoms. One specimen, $ [H 929]. Same locality. 30th May 1903. 9-10 
fathoms. Temperature about 28°. One specimen, $ [H 936]. 





Fig. 6. — Hectocotylised arms of Moschites charcoti. a, oral aspect of the 
extremity. Natural size. 

Previous Records. — Booth-Wandel Island. Lat. 65° 05' S. Among algae on the 
beach. 3rd September 1904. 

The hectocotylised arm (fig. 6) is short and stout ; the ridge bounding the 
seminal groove is very well marked, and is continuous with the margin of the 
umbrella. The groove itself is broad and deep, the extremity measures 7x5 mm. ; 
the longitudinal groove is triangular in form, and has four transverse ridges in its 
bottom. 



280 



DR WILLIAM EVANS HOYLE ON THE 



The radula is shown in fig. 7. 

According to coloured drawings made on the Expedition, the male of this species 
is dull stone colour above, deepening to brown in the centre of the back ; the 




Fig. 7. — Radula of Moschites charcoti, 6 [H 924]. x 50. 

female is much paler, with a pinkish tinge above, almost white below. The 
colours would, however, probably undergo change according to the varying state 
of contraction of the chromatophores. 

Sepiolidje. 

Euprymna sp. 

Locality.— Station 482, Saldanha Bay, Cape Colony. 19th May 1904. 8-10 
fathoms. Trawled. One specimen, too young to determine [H 934]. 

■Sepiolid gen. et sp. ? 

Locality. — Entrance to Saldanha Bay, Cape of Good Hope. 21st May 1904. 
25 fathoms. 

A head and arms, much macerated [H 1367]. 



Loliginid^e. 
Loligo reynaudi, d'Orbigny, 1845. 

Locality. — Station 480, eight miles north of Dassen Island, Cape Colony. 35 fathoms. 
Between 2 and 2.30 p.m., 18th May 1904. One specimen, ? [H 927]. 

Twenty-six young specimens, thirteen $, twelve $, one damaged [H 930]. One 
somewhat damaged specimen, %, probably of this species [H 1372]. 

Previous Records. — Cape of Good Hope ; False Bay, Cape Town. 

It is quite possible that some of the young specimens recorded as females 
may be males in which the secondary sexual characters were as yet undeveloped. 



CEPHALOPODA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 281 

Sepiid^e. 
Sepia australis, Quoy and Gaimard, 1832. 

Locality. — Station 480, eight miles north of Dassen Island, Cape Colony. 35 
fathoms. 18th May 1904. One specimen, $ [H 932]. 

Previous Records. — Cape of Good Hope, Agulhas Bank ; North Queensland ; 
New South Wales. 

This is not S. australis, d'Orbigny : that author changed Quoy's name to S. capensis, 
and gave the name S. australis to a quite different form. 

The tentacular club (fig. 8) shows three suckers much larger than the others, 
which diminish in size towards the tip, the third being about half the diameter of 
the first. 




Fig. 8.— Tentacular club of Sepia australis [H 932]. x7*5. 



Hemisepius typicus, Steenstrup, 1875. 

Locality.— Station 482, Saldanha Bay, Cape Colony. 19th May 1904. 
fathoms; trawled. Two specimens, ? [H 933 and 1380]. 
Previous Record. — Table Bay, Cape Town. 



8-10 



OnYCHOTEUTHID/E. 

Onychoteuthis ingens, Smith, 1881. 

Locality.— OR the South Orkney Islands. Lat. 60° 10' S., long. 42° 35' W. 
6th February 1903. From the stomach of a Koss' seal: a number of half-digested 
fragments [H 925]. 

Station 325, Scotia Bay, South Orkneys. 1st January 1904. One specimen [H 928]. 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 14). 43 



282 



DR WILLIAM EVANS HOYLE ON THE 



A drawing of the radula is given in fig. 9, as its form differs in a few details from 
that shown by Smith. # 




Fig. 9. — Radula of Onychoteuthis ingcns [H 925]. x 25. 
HlSTIOTE UTHID^E. 

Histioteuthis sp. juv. 

Locality.— Station 468, South Atlantic. Lat. 39° 48' S., long. 2° 33' E. 29th 
April 1904. 2645 fathoms. One specimen [H 940]. 

The specimen is somewhat damaged. The interbrachial membrane is slightly devel- 
oped. One arm shows the pigmented organ at the extremity, which, so far as I am aware, 
is characteristic (in this family) of the genus Histioteuthis, although it is not alluded 
to in the diagnosis either of Pfeffer or Chun. In many respects it resembles the 
Challenger specimen called Histiopsis atlantica, which was also from the same region, 
but is pale and semi-transparent, whilst that was opaque and dull reddish in colour. 

Bathyteuthid^e. 

Bathyteuthis abyssicola, Hoyle, 1885. 
Benthoteuthis megalops, Chuu, " Cephalopoden," Wiss. Ergebn. deutsch. Tie/see Exped., p. 185, pis. xxiv.-xxvii. 

Locality.— Station 416, off Coats Land. Lat. 71° 22' S., long. 18° 15' W. 
Surface to 2300 fathoms. 1 7th March 1904. One specimen [H938], 

Previous Records. — Southern Ocean, lat. 46° 16' S., long. 48° 27' E. ; off 
Martha's Vineyard, U.S.A.; off Cape Mala, Gulf of Panama; off Cape Agulhas; 
Equatorial Indian Ocean. 

Professor Chun has adopted Verrill's name Benthoteuthis megalops for this 
species, on the ground that "sheet 50 of the Trans. Connect. Acad., vol. vi., in 
which Verrill's description is contained, bears (p. 399) the note 'April 1885." 
If my friend is content to accept this method of determining dates of publication, 
he may turn to sheet 34 of the Narrative of the Challenge?* Expedition, vol. i., 
first part, in which Hoyle's description is contained, and he will find that it bears 

* Proc. Zool. Soc, 1881, pi. iii., fig. 1 b. 



CEPHALOPODA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 283 

(p. 265) the date "1884." I was fully aware of both these dates when 1 prepared 
the Report on the "Challenger" Cephalopoda, but as a matter of fact neither of 
them is a date of publication. Sheet 50 of Verrill's Third Catalogue of Mollusca 
. . . of the Neiv England Coast was not published by itself, but along with sheets 
51-56, in a wrapper which bears the words, " Newhaven, April to June 1885." 
Therefore, under the most favourable construction, it cannot possibly have appeared 
before June, and careful inquiries which I made at the time led me to the conclusion 
that it did not make its appearance till July. I may further add that in the 
twenty-five years which have elapsed since the statement was published its accuracy 
has never been impugned. 

Cranchiid^;. 

Galiteuthis suhmi, Hoyle, 1886. 
? Procalistes suhmii, Lankester, Quart. Journ. Micr. Sci., vol. xxiv. p. 311, 1884. 
Taonius suhmi, Hoyle, Ceph. Challenger Exped., p. 192, pi. xxxii. figs. 5-11, 1886. 
Taonidium suhmi, Pfeffer, Synopsis Oegopsid. Ceph., p. 192, 1900. 
Galiteuthis armata, Joubin, Ann. Sci. Nat, (Zool.), se>. 8, vol. vi. p. 279, 1898. 
Galiteuthis (Taonidium) suhmii, Chun, " Cephalopoden," Wiss. Ergebn. deutsch. Tie/see Exped., 
p. 382, pi. lix., 1910. 

Locality.— Station 422, Weddell Sea. Lat. 68° 32' S., long. 12° 49' W. 23rd 
March 1904. Vertical net ; surface to 600 fathoms. One specimen [H935]. 

Previous Records. — South of Australia, lat. 47° 25' S., long. 130° 22' E. ; 
Mediterranean ; Equatorial Atlantic in the Guinea Current. 

This specimen has a mantle length of 45 mm., and is, therefore, considerably 
larger than that described by Chun (34 mm.) ; but nevertheless I could find no 
trace of the modification of the tentacular suckers into hooks as depicted by him 
{pi. lix. figs. 6, 7) ; still, the other characters agree so well that I have no doubt 
that it belongs to the same species as his. 

If it could be proved satisfactorily that the embryo described by Lankester 
really belonged to this species, his name would take precedence ; but at present 
it seems advisable to keep the name Procalistes suhmii for it, and to call the 
more mature specimens by the name adopted by Chun. 

Desmoteuthis sp. 
Locality. — Station 98, off Eio Grande, South America. Lat. 34° 2' S., long. 
49° 7' W. 28th December 1902. Mantle and fin, cast up by a petrel. Too 
fragmentary to determine. [H 1368.] 

I have not thought it necessary to encumber this Report with full bibliographical 
references ; these will be found in my Catalogue of Recent Cephalopoda and its two 
Supplements.* The drawings have been made by Miss I. M. Davenport, B.Sc, under my 
supervision. 

* Proe. Roy. Phys. Soc. Edin., vol. ix., 1886 ; vol. xiii., 1897 ; vol. xvii., 1909. 

National Museum op Wales, Cardiff, 
6th January 1912. 



( 285 ) 



XV. — On Branchiura sowerbyi Beddard, and on a new species of Limnodrilus with 
distinctive characters. By J. Stephenson, M.B., D.Sc. (Loncl.), Major, Indian 
Medical Service ; Professor of Biology, Government College, Lahore. Communi- 
cated by Professor Ewart. (With Two Plates.) 

(MS. received October 25, 1911. Read December 4, 1911. Issued separately May 23, 1912.) 

On a recent occasion, looking into a small shallow pool near Lahore, I saw here and 
there on its bottom a reddish appearance, which a closer examination showed to be due 
to innumerable worms, implanted by one extremity in the mud, and waving their free 
ends in the water. These were the Limnodrilus described below. A quantity of the 
mud taken for further examination yielded, though in very much smaller numbers, 
three other Oligochaetes, viz. Branchiura sowerbyi, Lahoria hortensis, and a species 
of Dero. 

Branchiura sowerbyi was originally described by Beddard (2) from specimens 
found in mud from the " Victoria regia tank " in the Royal Botanical Society's Gardens, 
Regent's Park. Its most distinctive feature is the possession of a double series of 
finger-like gills on the posterior part of the body, one row being placed along the mid- 
dorsal, the other along the mid-ventral line. Lahoria hortensis was originally described 
by me (9) from a small pond in the Lawrence Gardens at Lahore ; it is not common ; 
and, not having met with it for two years, I had begun to think that it had disappeared. 
It belongs to the Naididae ; like Branchiura sowerbyi, it possesses two rows of gill- 
processes, but these are either limited to, or best developed on, the anterior part of the 
body, and are dorso-lateral in position. Dero is well known as a Naid which possesses 
a small number of gill-processes at the posterior end of the body round the anus. 
Limnodrilus is remarkable in possessing a cutaneous capillary plexus in the posterior 
part of its body, — better marked in this, apparently, than in most of the other species 
of the genus ; and this feature, and the constant waving movements of its tail, have 
doubtless a respiratory significance. 

There were here, therefore, living together, four genera of Oligochaetes, with peculiar 
and specialised respiratory arrangements ; and, by a curious coincidence, three of them 
were among the very few known Oligochaetes which possess gills. 

The coincidence, 'however, does not end here. Beddard, in his paper on Branchiura 
sowerbyi, relates that the mud which provided him with this form also contained three 
or four examples of Chxtobranchus semjjeri, a gilled Naid previously described from 
Madras by Bourne (4). Branchiura sowerbyi has been recorded twice since then 
(1892) (v. inf.), but Chsetobranckus has apparently not been seen again. Lahoria, 
however, resembles Chsetobranchus very closely in many ways,* the chief difference 

* Michaelsen (7) indeed includes it in the genus Branchiodrilm, Mchlsn. ( = Ghait> bronchus, Bourne). I have 
discussed the question in my original paper (9) ; the point depends on the value to be attached to " cephalisation,'' 
TRANS. ROY. SOC. ED1N., VOL. XLVIII. PART II. (NO. 15). 44 



286 DR J. STEPHENSON ON 

being the absence in Lahoria of gills and dorsal setae on the segments from the second 
to the fifth, and their presence on these segments in Chsetobranchus. It is thus note- 
worthy that the coexistence of Branchiura soiverbyi and Chsetobranchvs in artificial 
surroundings in England should be paralleled by the coexistence of Branchiura 
soiverbyi and Lahoria under natural conditions in India. 

The pond in which the worms were found was one of a series occurring in a small 
nullah on the outskirts of Lahore. The surface drainage of the jail runs into this 
nullah, so that, unlike most small ponds in this neighbourhood, it is probably seldom 
altogether dry. At times, e.g. after rain, there is a continuous stream of water along 
the nullah. 

Methods.— 'Hie best method for the observation of the living animals is the employ- 
ment of the binocular microscope with comparatively low powers ; this is the easiest 
way of obtaining a conception of the complicated relationships of the blood-vessels, and 
the mode of their contraction. 

The setae are best studied by killing the animals by a narcotic (methyl alcohol, 
chloretone) and leaving them in the water for a few hours till they become soft and are 
just beginning to disintegrate. A worm is then carefully placed on a slide in glycerin 
and the coverglass lowered ; the weight of the coverglass is sufficient to cause a 
specimen in this condition to flatten out completely, and the setse, retaining their 
arrangement, can then be viewed in one plane. 

Parts of the genital system, and especially the chitinous penis sheath of the 
Limnodrilus, can be isolated by teasing under the binocular microscope. 

Serial sections (5-10^) were stained with Heidenhain's iron-hsematoxylin followed 
by eosin, and with Delafield's htematoxylin. 



Branchiura soiverbyi Bedd. 

Since the original discovery of this worm by Beddard it has been recorded by \j. 
Perrier (8), who has found it in different years at several places in the Rhone at Touron 
(Ardeche), and by Michaelsen (6) in a warm-water tank of the Botanical Gardens at 
Hamburg. Perrier does not give any anatomical description of his specimens, while 
.Michaelsen states that, apart from the genital apparatus, which he describes in detail. 
his specimens agree essentially with Beddard's. 

I do not propose to give a complete description of the worm. My own specimens 

i.e. the differentiation of a certain number of the anterior segments of the body, — especially their differentiation by 
the absence from them of the dorsal scries of setae. The presence or absence of this cephalisation has hitherto, I think, 
always been regarded as of generic importance, and the extent anteriorly of the dorsal setaj is always mentioned in the 
generic diagnoses of the Naididaj ; and so Bourne, Beddard (3), and Michaelsen (5) all give the distribution of the 
dorsal seta: (as far forwards as the second segment) as a generic character of Chactobranchun. 

It is, 1 think, very possible, as 1 have myself suggested, that cephalisation will lose much of the importance 
hitherto assigned to it. But until the question has been further discussed, and until authorities are agreed with 
respect to this far-reaching alteration in our ideas as to specific and generic characters in the Naididai, it seems better 
(o abide by the more general opinion on this point. 



BRANCHIURA S0WERBY1 BEDDARD. 287 

were none of them mature sexually, so that a comparison with Beddard's and 
Michaelsen's specimens is in this respect impossible. I wish, however, to refer in some 
detail to the setse, the gills, the body-wall, the ccelom and its partitions, the circulatory 
and nervous systems ; and to bring out certain points which are new, or in which my 
specimens appear to differ from those described by Beddard. 

Seven specimens of this worm were obtained. In length they were, when extended, 
two inches or less, i.e. considerably longer than those examined by Beddard ; one, much 
the smallest, measured only about two-thirds of an inch. They were fairly stout, in 
breadth about a millimetre or more, very contractile, with both ends tapering to a blunt 
point. In colour they were a pinkish grey, with whiter and more translucent margins. 

The prostomium is bluntly conical ; the number of segments 74-116. 

The setse are of three kinds, single and double -pointed needles, and hair-setae ; the 
needles of both forms occur in both dorsal and ventral bundles, the hair-setee only in 
the dorsal. 

The needle-setse (fig. 1) are mostly forked in the anterior part of the body, single- 
pointed needles being relatively few ; in the posterior bundles the reverse is the case, 
the single-pointed being here the more numerous. In the double-pointed setse the 
outer point, i.e. that on the outer side of the curve of the shaft, is the smaller; inter- 
mediate forms between the single- and double-pointed setse are met with, in which the 
outer point is still smaller, or, it may be, scarcely recognisable. The single-pointed 
setse and the intermediate forms are not produced from the forked setse by wear, 
since, especially in the posterior part of the bod)', the formation of new setse may be 
observed, and many of these have single points from the beginning. In length these 
needle-setse are about 120m, in breadth 6-7m ; they have the usual double curve; the 
nodulus, not mentioned by Beddard, is distal to the middle of the shaft (distal to 
nodulus : proximal to nodulus : : 2 : 3) ; and the seta ends internally by tapering 
somewhat, not as shown by Beddard in a broad square end. 

The hair-setse are a little longer (180-164/x) and much slenderer than the needles ; 
they are straight, and show no other distinctive marks ; they are confined to the dorsal 
bundles of the anterior part of the body, being absent from the whole of the gill-region ; 
there is never more than one in each setal bundle. 

The setal bundles begin both dorsally and ventrally in the second segment, and 
cease some distance in front of the hinder end of the animal, the last ten segments or so 
being devoid of setse. There may be as many as six needle-setse in a ventral bundle ; 
five, four, or fewer are also met with. In the dorsal bundles the most that I have 
observed is five needles and one hair- seta. 

The gills correspond in their appearance to the description given by Beddard. 
They are not ciliated, are situated along the mid-dorsal and mid-ventral lines, and 
occupy the posterior fourth to two-fifths of the animal's length. There were from 38 
to 55 pairs, on a corresponding number of segments. In length they are about equal 
to the diameter of the body, but they diminish in size towards the anterior, and less 



288 DR J. STEPHENSON ON 

markedly towards their posterior limit ; anteriorly, before fading away completely, they 
become mere tubercles. They are in constant movement, in some degree reminding the 
observer of cilia ; the resemblance is frequently increased by the regular propagation of 
the beat along the line of the gills, in a postero-anterior direction, like the propagation 
of a ciliary wave ; but there is this difference, that the movement of the individual gills is 
a swinging movement from side to side, i.e. in a direction at right angles to the direction 
of propagation of the wave, not, as in the case of ciliary movement, in the same direction 
as the wave. The structure of the gills will be described with that of the body-wall. 

In addition to the oscillations of the gills themselves, the whole posterior part of the 
body of the worm is continually performing undulatory movements. 

The body-wall (figs. 2-5) consists of the usual layers. The surface epithelium is 
high and columnar over the body generally, low over the gills (fig. 6). The subjacent 
circular muscular layer is thin, and extends unbroken throughout the body. The longi- 
tudinal muscular layer, on the other hand, is thick, but is characterised by inequalities 
of distribution ; in the anterior part of the body (fig. 2) it is of the same thickness at 
all parts of the body-circumference, but in the gill region inequalities appear, while 
posteriorly (fig. 3) it is absent from parts of the body-wall. The position of the bands of 
fibres that persist varies : they may be lateral, one on each side ; or one dorsal and one 
ventral ; or there may be one, a ventro-lateral, band only. The position of the absent 
fibres is taken by an indefinite network with a few cells, the appearance being some- 
what that of a scaffolding or meshwork from which the fibres have dropped out. 

An inner circular layer of muscular fibres is well marked in the anterior part of the 
body (fig. 2). The muscular fibres of the septa are continuous with the fibres of this 
layer ; and special bundles of fibres, passing vertically between the sacs of the dorsal 
and ventral setse on the same side, may be considered as belonging to it (cf. also 
Lvmnodrilus socialis, post). 

In the gill region the circular muscular layer of the body-wall is continued over the 
gills without splitting, while the longitudinal layer closes the base of the space within 
the gill, the fibres maintaining their straight course without being deflected. In the 
interior of the gill, within its muscular layer, are a number of branched cells whose 
processes anastomose (a, fig. 6) ; these may entirely fill the distal part of the gill-cavity. 
Proximally, near the base of the gill, there is always a space ; cells and cell- processes 
may stretch across it, and other cells may form a definite, almost epithelial layer on the 
inner surface of the muscular coat (fig. 6). The two vessels of the gill run on opposite 
sides, within the muscular coat. 

Beddard, who describes a septum shutting off the cavity of the gill from the general 
body-cavity, nevertheless speaks of the gill-cavity as " evidently belonging to the 
coelom." I do not think this follows either from his description or mine; the gill- 
cavity appears to be a space between the muscular layers of the body-wall, which 
undergo a separation consequent on the circular layer alone being continued over the 
gill. Beddard also states that the movements of the branchiae are caused by muscular 



BRANCHIURA SOWERBYI BEDDARD. 289 

fibres, elongated, fusiform, with central nucleus, or sometimes branched and star-shaped, 
which traverse the cavity of the branchia from side to side. These fibres are obviously 
the branched cells which 1 have described above ; but I can find no reason to suppose 
that they are muscular, either in Beddard's description or my own preparations. On 
the one hand, the observed movements of the gills could not be accounted for by the 
contraction of such fibres; while, on the other hand, all the movements shown by the 
gills in my specimens could be caused by a contraction, on one or other side of the gill, 
or on both sides simultaneously, of the fibres derived from the circular muscular coat. 

A fairly obvious feature in the constitution of the body-wall is the lateral line. It 
exists on each side within the circular muscular layer as an aggregation of cells, which 
in its situation and in its relation to the longitudinal muscular layer recalls the similar 
structure in the Nematodes. The circular muscular layer is continuous over the line of 
cells, but the longitudinal layer is completely divided by it on each side (figs. 2, 3, 4). 
The cells are of somewhat small size, their outlines indistinct, their nuclei ovoid or spindle- 
shaped. The line extends, midway between dorsal and ventral rows of setae, from the 
third or fourth segment to within a few segments of the posterior end. It is easily 
followed in transverse sections ; its nature can also be seen in vertical sections where 
the lateral body-wall is cut tangentially ; the line of cells is seen to form a continuous 
track, of uniform width, in the substance of the longitudinal muscular layer. The line 
is continuous, however, only as far back as the anterior gill-region ; behind this it is 
interrupted, and consists of a series of segtnentally arranged groups of cells in the 
posterior part of each segment in front of the septum. Near the posterior end of the 
body the groups of cells, as seen in transverse sections, project further inwards and 
spread out somewhat in the body-cavity (fig. 4). The lateral line is also the situation 
from which numerous bundles of muscular fibres arise.* 

The ccelom and its partitions. — The partitions which separate off the cavities in the 
gills, and the question of the ccelomic nature of these latter, have already been considered. 

The septa are extremely thick and muscular in the anterior part of the body (fig. 2). 
The first septum is f ; from here onwards to -^o the septa have the above character ; \^ 
and all succeeding septa are thin ; they contain, however, muscular fibres, both radial 
and circular. 

Beddard states that "there appears always to be a partition which shuts off the 
upper part of the coelom from the lower part " (he is speaking of the branchial region) ; 
the upper cavity he found " exclusively occupied by the intestine, the lower cavity by 
the nervous system and the principal blood-vessels ; " in his figure the partition is shown 
as a thin peritoneal membrane, on which are a number of nuclei. 

I do not find any partition of this nature in my specimens. There are, however, in 
the branchial region a number of well-defined, fairly thick bands, which pass across 
from one lateral line to the other. These do not form a continuous sheet, but have a 

* For the most recent contribution on the subject of the lateral line, v. H. Pointner, " Beitriige zur Kenntnis der 
Oligochaetenfauna der Gevvasser von Graz," Zeit. f. wiss. Zool., Band xcviii., Heft 4, 1911. 



290 DR J. STEPHENSON ON 

segmental arrangement, being present for some distance in front of each septum (fig. 5). 
In this situation, moreover, the intestine and dorsal vessel are above, the nerve cord and 
ventral vessel below the band. 

Other muscular strands are also to be seen (cf. fig. 5) : thus fibres may be observed 
passing from the transverse bands in an oblique direction downwards to the body-wall. 
There are, in addition, a number of strands in the upper part of the body-cavity which 
pass between the intestine and the parietes ; these appear to be each constituted by a 
single cell, with a large ovoid nucleus situated to one side of the strand. 

In the most posterior part of the body the ventral portion of the coelom may be 
completely filled by the blood-vessels, nerve cord, and a number of cells and fibres, so 
that there is here no free space below the level of the transverse bands (fig. 5). 

Circulator)/ system. — The dorsal vessel is, as described by Beddard, not dorsal at 
all for the greater part of its course ; it is situated ventrally, or rather ventro-laterally 
on the left side (figs. 3, 8), or it may be in places directly on the left of the 
intestine (fig. 5). Though not lying immediately on the intestinal wall, it has 
nevertheless a closer relation to the gut and to the chloragogen cells than has the 
ventral vessel, and the branches which unite it with the plexus on the intestine are much 
shorter than those which connect the plexus with the ventral vessel (fig. 8). In sections 
it is as a rule larger than the ventral vessel. It passes up the left side of the 
alimentary canal in the eleventh and tenth segments, and thenceforward is dorsal in 
position. In life it is contractile. 

The ventral vessel (figs. 3, 5) is on the right of the dorsal in the posterior part of 
the body ; it lies on or near the nerve cord ; it is connected by numerous vessels with 
the intestinal plexus. Anteriorly it is formed by the union of the hearts. Its place 
is taken in front of the ninth segment by a trunk which is formed in segment v by 
the union of two vessels coming from the anterior end of the body, and presumably 
originating in the forking of the front end of the dorsal vessel ; this ' anterior ventral ' 
vessel receives the lateral loop vessels on each side, including the large loops in 
segment viii from the supra-intestinal ; very soon thereafter it becomes enveloped in 
the chloragogen cells and disappears on the ventral surface of the alimentary canal. 

There is a rich intestinal plexus in the gut-wall, which in segments x-xiii becomes 
almost a sinus ; sections through this region frequently show a continuous blood-space 
all round the circumference of the alimentary tube. Besides the connections with 
dorsal and ventral vessels, the plexus is joined to the supra-intestinal by a series of 
wide communications. 

Owing to the comparatively large size and consequent opacity of my specimens 
during life, it was not easy to trace out in the living worms all the details of the course 
of the vessels to the gills and body-wall. Fig. 7 represents what could be seen ; and 
in addition I have worked out the point in sections (fig. 8). In the gill-region the 
dorsal vessel gives in each segment two branches, one dorsalwards, on the inner surface 
of the body-wall (d.d.) } which gives a twig across the mid-dorsal line and then enters 



BRANCH1URA SOWERBYI BEDDARD. 291 

the dorsal gill ; and one ventralwards (n.d.), with a similar course and distribution on 
the ventral side. The two branches may, in the region of the anterior gills, arise in 
common (fig. 7). The ventral vessel gives two branches : one goes to the right and 
bifurcates almost immediately into two divisions, which correspond to the two branches 
of the dorsal vessel, — one running dorsalwards on the inside of the body-wall to the 
dorsal gill (d.lr.v.), the other taking a short course ventralwards to the ventral gill 
(v.br.v.). The second branch of the ventral vessel (l.br.v.), leaving the main trunk 
on the left side of the latter, arches over the nerve cord and goes to the left side of 
the body-wall ; it gives a branch to the right below the nerve cord, which seems, 
according to sections, sometimes at least to meet and anastomose with the branch 
from the ventral vessel to the ventral gill ; thus in these cases a ring is formed round 
the ventral nerve cord. 

I am not satisfied that the arrangement above described is constant ; in one series 
of sections the two vessels going to the dorsal gill were derived, one from the dorsal 
vessel in the usual way, described above, and the other from the branch from the 
ventral vessel to the left side of the body- wall (i.e. the dorsally directed branch of 
the vessel l.br.v. in fig. 8 went up into the gill). 

In front of the gills the parietal vessels have the form of lateral loops, two per 
segment, — one on the anterior face of the septum, the other more anteriorly in the 
segment. There are also, as described by Beddard, a number of longitudinal vessels, 
some of which connect successive loops, while others, smaller, pass over several 
segments at least. The muscular layers of the body-wall are penetrated by numerous 
twigs, but no capillaries seem ever to enter into the surface epithelium. 

The supra-intestinal vessel is hardly or not at all distinguishable in front of the 
sixth segment, and it is small as far back as segment viii ; here two considerable loops, 
one on each side, put it in connection with the 'anterior ventral' vessel ; these loops, 
though not as large as the hearts in the next segment, are still extremely conspicuous 
structures. In segment ix it gives rise to the first pair of hearts, and in the same 
segment a number of vessels radiate from it across the body-cavity to the parietes. 
In this part of its extent it lies directly beneath the dorsal vessel, and, like the latter, 
gets round to the left side in the posterior part of its course. It is largest in segments 
x and xi ; it communicates by wide channels with the intestinal plexus, and is 
throughout in close relation with the gut and its chloragogen cells. It becomes 
smaller in segment xii and very soon disappears. 

The hearts are two pairs, in segments ix and x. The first pair originate above 
from the supra-intestinal vessel, pass downwards close to the gut, then, taking a 
backward course, leave the intestine and perforate septum ^ separately ; continuing 
parallel for a short distance, they then unite in x to form the ventral vessel. The 
second pair appear to be more tortuous and of rather smaller calibre ; they originate 
above from the dorsal vessel, which is here on the left side, so that the heart of the 
right side arches over the intestine ; they pierce septum \^ separately (this septum 



292 DR J. STEPHENSON ON 

thus transmits three vessels ventral to the intestine) and unite with the ventral vessel 
iu xi. Where the hearts join in the ventral vessel there are in both segments (x and 
xi) projections of cells into the lumen ; these may, according to the evidence of 
sections, entirely block the cavity of the vessel. 

Lastly, the anterior loops occur in segments ii-viii ; they are non-contractile, 
and the posterior are larger than the anterior ; in vii they are of about the same size 
as the dorsal or the ventral vessel, and in viii they are so large as to resemble the 
hearts. In ii-vii they run from the dorsal vessel to the anterior ventral, or to the 
branches which unite to form the latter ; in viii, as already said, they run from the 
supra-intestinal to the anterior ventral. They give branches to the body-wall, which 
in vi, vii, and viii are of moderately large size. It may be added that the dorsal 
vessel also gives considerable branches to the body-wall in vi and vii, and the parietal 
vessels form a plexus in the prostomium and most anterior segments. Behind segment 
iv the parietal vessels have a longitudinal course, running parallel and fairly close 
together, so that about 20 such vessels are visible on examining the dorsal surface of 
the worm ; there would thus be about 40 longitudinal vessels in all at any particular 
level, each of which runs through several segments without losing its individuality ; 
this arrangement, as before noticed, ceases at the gill region. 

The only other feature which I propose to select for special mention is the presence 
of remarkable giant fibres in the ventral nerve cord. They may be seen throughout 
the body on the dorsal side of the cord. They are of different sizes ; the larger appear 
as tubes in transverse section, with a roughly circular or oval outline ; a part of 
their lumen is always empty, and a part, to one side, is occupied by a substance which 
sometimes has the appearance of a fairly solid mass, sometimes that of a thin 
coagulum ; the shape of this contained matter varies, being sometimes ovoid, sometime.-; 
quite irregular ; at times it is connected here and there with the wall of the tube by 
a number of branching threadlike extensions ; sometimes it occupies a considerable 
portion, sometimes very little of the lumen (figs. 3, 5, 6). Other fibres, also running 
longitudinally on the dorsal and dorso-lateral surface of the cord, are not tubular, and 
in their staining reactions somewhat resemble muscle fibres. 

The number of these large fibres varies from place to place ; there are in all, in any 
section, about five to ten ; they are most often six or seven in number, of which one to 
three have the tubular appearance described above, the rest being larger or smaller 
solid fibres. 

The actual size of the tubular fibres also varies. One is generally larger than the 
others, — in the posterior part of the body very much larger, — and this one is there 
constantly to the left of the middle line. These fibres are on the whole larger towards 
the hinder end of the animal ; in the anterior segments they average about 20m in 
diameter ; as the gill region is approached they may be 40m ; while posteriorly one, but 
ordy one at any given level, may be as much as 71m ; these large fibres are, however, 
much constricted at the septa in this region, e.g. to about 15m or even much less. 



BRANCHIURA S0WERBY1 BEDDARD. 293 

The fibres are thus very conspicuous in sections through the posterior part of the 
body, and especially so is the largest fibre of the group (fig. 3), not only because of its 
actual size, but because, the whole section being so much smaller, it occupies relatively 
a far larger space than is the case anteriorly ; its diameter may be nearly as great as 
that of the whole nerve cord, and may actually be greater than that of the ventral 
vessel close to it. 

I am inclined to think that these fibres, or at any rate the specially large one, were 
seen by Beddard and described by him as the dorsal vessel. As will be seen, his 
description of the dorsal vessel is altogether at variance with the condition found by 
me, while it corresponds exactly to that of the large giant fibre. 

The dorsal vessel is described by Beddard as being below the partition dividing 
the ccelom into upper and lower parts; as having thicker walls, and much less blood in 
the lumen, in sections, than the ventral; the blood in the dorsal vessel is, he states, 
never so darkly stained by carmine as the blood in the ventral vessel, to explain which 
he supposes that possibly the muscular walls of the dorsal vessel are particularly im- 
permeable to the staining fluid ; when fully expanded the dorsal is stated to be of about 
the same calibre as the ventral vessel, but in certain parts its lumen was so contracted 
that the vessel could only with difficulty be recognised ; the openings of the branchial 
vessels into the dorsal vessel were not seen, since at the point where these vessels 
should open it always happened that the dorsal vessel was very much contracted, while 
the end of the branchial vessel was much dilated (i.e. the dorsal vessel, according to 
Beddaed's interpretation, would thus have a regularly moniliform shape, being con- 
tracted almost to obliteration once in each segment). 

From my description, however, it appears that the dorsal vessel is above the incom- 
plete muscular partition of the coelom, and has no close relation to the ventral nerve 
cord ; its lumen is in my specimens always full of blood, which has the same staining 
reactions as blood elsewhere (this, of course, is only what would be expected in sections 
stained on the slide) ; I found the dorsal to be often, if not usually, of greater calibre 
than the ventral vessel ; the openings of the vessels into it were always well marked 
and patent, and there were no constrictions of importance along its course. 

The specially large giant fibre, on the other hand, is below the ccelomic partition, 

and has the relation to the nerve cord described by Beddard for the dorsal vessel (on 

the left side just above the cord) ; the contents of the tube, which usually have the 

appearance of a coagulum, do in fact stain differently from the blood in the vessels ; in 

diameter it is seldom (and only at the posterior end) as large as, or larger, than the 

ventral vessel ; and at the septa it is regularly much constricted, so as to be, in the 

posterior portion of the animal, moniliform ; while, of course, no vessels are to be seen 

opening into it. # 

* The appearance of giant fibres lias led other observers also to consider them as tubes containing a coagulable sub- 
stance, e.g. Semper, who also found that the coagulum reacted to stains quite differently from blood-plasma ; in the 
crayfish the fibres have been held to be blood-vessels. Compare Ashworth (1), section ii., " Historical Account of the 
Giant Cells and Giant Fibres of Annelids." 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 15). 45 



294 DR J. STEPHENSON ON 

I have only a few remarks to make on the other systems. The alimentary canal 
is a fairly uniform tube throughout, showing little differentiation into distinct regions. 
The pharyngeal epithelium is markedly ciliated ; chloragogen cells begin in segment vi, 
and cease in the anterior gill region ; the canal becomes somewhat wider in segment xi, 
from which point it may be spoken of as intestine ; the anus is dorsal. The ' septal 
glands,' in segments iii, iv, and v, are collections of cells, not massed together in 
definite lobes, but surrounding the alimentary tube on all sides in the posterior part of 
each of the three segments ; the septa are here bulged backwards, and the funnel-shaped 
space so formed is filled with the cells (fig. 2, p). The glands are thus, apparently, 
merely collections of peritoneal cells. 

The nephridia are long, closely coiled tubes, with a pear-shaped reservoir near the 
external aperture ; the reservoir is less marked in the posterior nephridia. They begin 
in segment xii, and cease from fifteen to thirty segments in front of the posterior end ; 
the external apertures are in line with, and in front of, the ventral setal bundles, near 
the anterior margin of each segment. 

The cerebral ganglion is deeply indented anteriorly, less so behind ; the dorsal 
vessel, here divided, is closely applied to its under surface. The ventral nerve cord is 
relatively very large in the posterior part of the body (fig. 5), and may be equal in 
diameter to the intestine ; it is even absolutely larger than in the (much thicker) 
anterior part of the body ; thus in one specimen, when sectioned, the transverse 
diameter of the cord was 61u anteriorly, 110^ where the gills commenced, and 82m in 
the posterior gill region. 

Unfortunately none of my specimens were sexually mature. Testes were present 
in x, and ovaries in xi, and in one specimen the male deferent apparatus was beginning 
to form, in the shape of a funnel on septum ^, while a backward pouching of the same 
septum indicated the commencement of the vesicula seminalis ; but there was nothing 
distinctive to be discovered. 

Limnodrilus socialis, sp. nov. 

The mode of occurrence of this worm has already been described. On a subsequent 
occasion I found it even more abundantly in the same locality ; the small pools were 
beginning to dry up, and the water was everywhere very foul ; the worms occurred in 
large tangled masses of sometimes several pounds in weight, their tails waving as 
before ; a considerable proportion were sexually mature on both occasions (December 
and February). Again, on March 1st, there were still numbers of the worms in the mud 
of the nullah ; a quantity of these were taken, and on examining them I was surprised to 
find that the large majority were headless, the whole of the anterior segments, including 
all the genital organs, having disappeared. Thus, in each of two batches of over fifty 
individuals, only four complete specimens were found. The worms nevertheless behaved 
us usual, waving their posterior ends, and contracting on being disturbed; when isolated 



BRANGHIURA SOWERBYI BEDDARD. 295 

in a dish they appeared to move and coil themselves up in the same way as perfect 
individuals (see below). 

A possible explanation of this curious circumstance seems to be that by the expulsion 
of the genital products, which apparently takes place towards the end of February, the 
anterior segments of the body are so much damaged that they die and are thrown off ; 
the worms, however, continue to live, though it may be doubted whether they are 
capable of regenerating the anterior end ; probably they die after a time, and the 
whole generation thus perishes each year. 

On touching any part of a mass of these worms they cease their waving movements ; 
a feeble disturbance merely causes them to hold themselves rigid, while a more violent 
one causes a general contraction of all the individuals. Contraction also takes place if 
the mud near them be disturbed. It looks, at first sight, remarkably easy to scoop up 
a number of the worms ; but as soon as the spoon touches the mud the waving tails 
vanish in a flash, and however quickly the scoop be made, probably few will be ob- 
tained. This, of course, is not the case with the large masses described above, which 
are so tightly intertwined that they cannot escape thus. The same waving movements 
and sudden contractions are seen in the worms kept in a vessel in the laboratory, where 
a slight jarring of the table is sufficient to arrest the movements. Isolated individuals 
usually coil themselves up into a spiral on being disturbed. 

External features. — The usual colour is a pale reddish brown, deeper anteriorly 
than posteriorly ; by reflected light under a low-power binocular microscope the posterior 
part of the body often has a patchy, opaquish yellow colour, which has the appearance of 
being due to a golden-bright granular deposit on the inner surface of the body-wall ; it 
is probably in reality a deposit in the peritoneal cells. Some specimens from the foul 
pools mentioned above were black in their posterior half ; and it may be mentioned, by 
way of comparison, that I have found Clitellio arenarius, usually of a red colour, 
completely black when inhabiting a part of the shore contaminated by sewage. 

The length of the worms when extended may be as much as three inches ; smaller 
specimens measure one and a half or two inches. Their greatest breadth is less than a 
millimetre. The number of segments is commonly about 110 ; there is a double annula- 
tion in the first few. The prostomium is bluntly conical. The clitellum includes the 
eleventh with more or less of the twelfth segment. 

The setse (fig. 9) of both dorsal and ventral bundles begin in segment ii, and are of 
the same form. They are moderately stout, have the usual double curve, and are bifid 
distally ; the proximal prong of the fork is shorter and stouter than the distal (about 
three-quarters as long, and one and a third times as thick at the base) ; the nodulus is 
distal to the middle of the length of the shaft (proximal : distal : : 3 : 2, or thereabouts). 
The length of the setye in the anterior part of the body is about 115m, but posteriorly it 
is less, the average being about 80m- The number of setse per bundle is six, seven, or 
eight in the anterior part of the body, diminishing to three or four posteriorly. 

In the first few segments the prongs of the fork may have a slightly different 



296 DR J. STEPHKNSON ON 

character, being longer and separated by a smaller angle than elsewhere. In the eleventh 
segment the ventral setse are absent, their position being occupied by the male 
aperture. 

The body-wall (figs. 10, 11) consists of the usual layers. In addition to the circular 
and longitudinal muscular layers, there is regularly present a well-marked muscular 
band, passing vertically through the body-cavity, between the dorsal and ventral setal 
sacs of the same side in each segment (fig. 10). The 'peritoneal cells lining the inner 
surface of the body-wall have a distinctive character (figs. 10, 11); they are fairly large, 
ovoid, and transparent ; in sections they are very slightly coloured, the only part of the 
cell-body which takes up the stain being a quantity of contained granular matter ; 
the circular or oval nucleus, with a well-marked nucleolus, is conspicuous ; they resemble 
chloragogen cells, but are without the numerous yellow particles which characterise these 
latter. Similar but larger cells, pyriform in shape, attached by their stalk-like narrow 
ends, are also found on septa f and -|-, and round the nephridia of segments vii and viii. 
There is a lateral line (fig. 10), similar to that described for Branchiura sowerbyi 
(see p. 289) ; the cells composing it extend peripherally outwards as far as the circular 
muscular coat, and divide the longitudinal layer along a line which runs midway between 
dorsal and ventral setal bundles ; in this worm, however, the lateral line is well marked 
only in the anterior segments. No free caelomic corpuscles were seen, except a few 
yellow granular cells, which were probably detached chloragogen cells. 

Alimentary canal. — The pharynx extends backwards through the third segment ; 
its tall columnar epithelium, as also that of the oesophagus, is ciliated. From the third 
to the seventh segment the ventral wall of the canal is raised into a prominent longi- 
tudinal ridge, due to the greater height of the epithelium along this tract (fig. II); the 
lumen is thus crescentic. Chloragogen cells begin in segment v. 

As appendages of the alimentary canal may be mentioned paired masses of cells in 
segments vi and vii. These are situated in the anterior part of the segment, ventro- 
laterally to the oesophagus ; in transverse section they have a pyriform shape, with tlie 
small end below, near the ventral vessel, and the glands of the same pair may unite 
with each other round the vessel (fig. 11). Some of the cells contain yellow granules 
like those of the chloragogen cells, but their general character is very different from 
that of these latter, inasmuch as they are more compact, more irregular in size and 
shape, and stain more deeply in both cell-body and nucleus ; moreover, though they are 
in contact with the chloragogen cells, they do not merge into them, and the masses 
have a distinct outline of their own. Similar but smaller aggregations were also found 
in segments v and viii. 

The circulatory system (figs. 12, 13, 14) is of considerable interest. The dorsal 
vessel is ventral in position for the greater part of its course ; it runs alongside and on 
the left of the ventral vessel ; both are sinuous, the ventral vessel more markedly so ; the 
convexities of their curves face away from each other. The dorsal vessel is somewhat 
laterally situated in segments xi and x ; it becomes more ventral again in ix, only to 



BRANOHIURA SOWERBYI BEDUARD. 297 

leave the ventral surface altogether and reach the dorsal side of the oesophagus in viii. 
The vessel is surrounded by chloragogen cells, and is situated in close contact with the 
intestinal wall, as far forwards as ix ; it then separates from the alimentary tube and 
becomes free in the body-cavity. It is contractile throughout its length. 

Beddard (8) mentions a ventral position of the dorsal vessel as a rare peculiarity 
among the Oligochaeta, which is found in the genera Branchiura and Dero. It is a 
curious coincidence that these two worms were found in association with the species 
under description, as related previously. 

The supra-intestinal vessel can be traced in the living animal as well as in sections 
as far forwards as the anterior part of segment v. It is covered by chloragogen cells 
throughout its course ; in the anterior part of its extent it appears in transverse sections 
as a fusiform space, a special channel of the gut plexus ; it is large in segment vii ; in 
segment viii it may, in the living specimen, be hardly visible (perhaps from accidental 
causes), or, on the other hand, it may be large and conspicuous ; in sections it is here 
larger than either the dorsal or the ventral vessel. 

In the anterior part of segment ix, situated transversely like a half ring on the 
dorsal side of the intestine, is a sinus-like blood-space, with the following connections 
(figs. 12, 13). Posteriorly it dissolves into a close network of small vessels in the 
intestinal wall. On the right side it is in open communication with a large vessel in 
the gut wall, which runs along the right side of the alimentary tube, gradually dying 
away, and becoming indistinguishable posteriorly at about segment xxi, while anteriorly 
it ceases as a distinct vessel in segment viii ; this, again, though a perfectly distinct, and 
indeed a very conspicuous, vessel in the living animal, is covered by chloragogen 
cells, and appears in sections as a special channel of the intestinal plexus. Anteriorly 
the sinus in segment ix may in favourable cases in the living worm be seen to be 
connected with the supra-intestinal vessel, though the channel of communication 
through septum f seems as a rule to be of inconsiderable width. The above descrip- 
tion is confirmed by sections ; the supra-intestinal can be traced as a well-marked 
vessel up to the septum, where all blood-channels are constricted ; on the other side 
of the septum there is no longer a supra-intestinal vessel, but a sinus encircling 
the upper half of the gut, from which, on the right side, the vessel described above 
takes origin. 

The intestinal plexus, which has already been referred to, extends throughout the 
length of the intestine, and reaches as far forwards as segment iv. For some distance 
behind the genital segments it is fed on the left side by a conspicuous series of branches, 
one in the anterior part of each segment, from the dorsal vessel ; and on the right side 
by twigs from the channel described above as running along the right side of the in- 
testine (fig. 14). 

The hearts (figs. 12, 13) are a single pair of vessels, which arise from the supra- 
intestinal anteriorly in viii, pass obliquely downwards and backwards, pierce septum | 
and gradually converge, to unite about the middle of segment ix. When full of blood 



298 DR J. STEPHENSON ON 

their anterior ends are much swollen ; their contractions are alternate, pass downwards 
and backwards, and have no time-relation to the contractions of the dorsal vessel. 

The genus Limnodrilus possesses in general two pairs of hearts ; the present species 
is therefore remarkable in having only one pair. There is, however, in the sexual animal 
a special loop to the genital organs, including the seminal vesicles and egg-sac. There 
is no trace of any such vessel in the non-sexual animal as a rule ; on one occasion, how- 
ever, in a specimen in which no genital organs were discoverable, a loop was seen in 
segment ix ; this specimen was perhaps just about to develop sexual organs. The loops 
develop along with the organs they supply, and soon extend backwards from segment 
ix in complicated windings ; in the fully developed sexual animal they may extend as 
far as xvi, — in one case two segments behind the posterior limits of the genital sacs, 
here appearing to be applied to the wall of the intestine ; they end ventrally in segment 
x by joining the ventral vessel (fig. 12). The loops are contractile, and it is a fascinating 
occupation to watch, the wave of contraction wandering along the complicated windings 
of the loop through segment after segment. It seems not impossible, therefore, to 
compare these genital loops with the second pair of hearts of other species of 
Limnodrilus ; they would be hearts developed only at sexual maturity, greatly increased 
in length, and modified for their special function. I have no actual note of their dorsal 
origin ; it is, however, presumably from the dorsal vessel, since the supra-intestinal does 
not exist behind segment viii. 

The ventral vessel (fig. 12) is formed anteriorly at about the level of septum i by 
the union of a pair of vessels, one on each side, coming from the front end of the 
animal. After receiving the lateral loops of segment vii it becomes smaller, and in 
the next part of its course it appears to be variable. In the living animal it was 
seen on one occasion to be continued as a fair-sized vessel to the junction of the hearts 
in ix, and so onwards after being thus reinforced. In three other cases it became 
very narrow ; but close observation at times when it was filled with blood showed 
that it was here also continued to join the angle between the hearts, or to one or 
other heart immediately in front of the junction. In other cases no connection could 
be made out, nor could any such channel be discovered in either of two series of 
transverse sections ; in these cases the vessel ended on the intestine in front of septum 
*, sometimes bending to the right, towards the anterior end of the channel in the 
right wall of the intestine (v. sup.) before disappearing. 

The condition therefore resembles in some degree that described in Bothrioneurum 
and Lophochseta (cf. Beddard, 3, p. 70), where the ventral vessel of the anterior 
segments is continued backwards on the intestine as a ' subintestinal ' vessel, while 
there is a fine channel of communication between this subintestinal vessel and the 
point <>f union of the hearts. The specimens in which the connection between anterior 
and posterior parts of the ventral vessel is altogether wanting show a condition similar 
to that described above for Branchiura sowerbyi ; and with this, again, may be compared 
Tubifex costatus {Heterochwta costata), Clap. (10), the only difference being that in 



BRANCHIURA SOWERBYI BEDDARD. 299 

this latter species the ' anterior ventral ' is continued backwards for a short distance 
on the intestine, beyond the level of union of the hearts. 

To return to the course of the ventral vessel in the species under description : the 
vessel formed in segment ix by the junction of the hearts extends to the posterior 
end of the body ; it is more sinuous than the dorsal vessel ; it is situated on the right 
of the nerve cord, the dorsal vessel being on the left. It is less intimately united 
to the intestine than the latter, not being covered by chloragogen cells ; it is somewhat 
widely separated from the intestine in the genital segments. 

The lateral commissures in front of the hearts form an elegant tracery of compli- 
cated loops. In the posterior part of the body the loops run on the anterior face of 
the septa, and give branches outwards to the body- wall, which form the cutaneous 
plexus. This extends through about the posterior half of the body ; it consists of 
a number of fair-sized vessels which penetrate the muscular coats, and so come to lie 
between the cells of the surface epithelium. There are four chief capillary vessels 
on each side in each segment, at about equal distances from each other ; the condition 
may therefore be compared with that in L. hojfmeisteri, as described by Vejdovsky 
(11, p. 116, and plate vii. figs. 16, 17); the special mode of branching there described 
is not, however, found in the present species ; there is no collection of chloragogen 
cells round the origin of the cutaneous branches ; and the cutaneous vessels in the 
form under description branch freely and anastomose. Though sometimes the secondary 
branches seem to come to a blind end, this is probably due to the pressure on the 
specimen, and I do not think they ever really end blindly. 

Nephridia are present in segments vii and viii ; there is then a hiatus as far as 
segment xiii. Thenceforwards, too, they are not present in every segment ; three or 
more consecutive segments may possess nephridia, and then they may be absent from 
one or two ; two or three more will have them, and so on ; but there is no general 
rule as to their distribution. The funnel is small, with long cilia round its margin ; 
these wave slowly and languidly ; but a few long flame-like flagella, arising from within 
the funnel and beating down the tube, are much more active ; these flame-like flagella 
are repeated several times in the course of the tube. The tube itself is long, loosely 
coiled, without a terminal vesicle ; it ends on the surface immediately in front of 
and lateral to the setal sac. The peculiarity of the nephridia of segments vii and viii 
— that they are surrounded by large pear-shaped peritoneal cells — has already been 
mentioned. 

The cerebral gan f/lion is deeply cleft in front, slightly so behind (fig. 15). 

The reproductive organs were well developed in a considerable number, perhaps 
in the majority of specimens. The testes are in x, attached by a narrow base to the 
junction of septum T 9 „ with the ventral body- wall. The funnel is also in x, on the 
opposite septum ; it has the usual characters. The vas deferens is in xi ; it is long 
and much coiled, wider in its first than in the later part of its course (39/x as against 
27m later) ; its lumen is also relatively and absolutely greater, and its walls stain 



300 DR J. STEPHENSON ON 

less deeply, in its first part. The cells of which its wall is composed have nuclei 
which are much elongated transversely to the axis of the tube ; their length may be 
as much as one-third of the circumference. The vas deferens is continued into the 
atrium, which begins in the posterior part of xi, the septum (-J4) being here much 
bulged backwards. The atrium is of an elongated pyriform shape, its first part being 
the broader (90m) ; the lumen is at first small, or even, in transverse sections, invisible ; 
only the basal portions of the cells take the stain. The penial end of the atrium is 
narrower, the lumen is more distinct, the whole of the cells are deeply stained so 
as to obscure the nuclei. The prostate is a fairly large, somewhat spherical cellular 
mass, continuous with the wall of the atrium, to the first part of which it is attached, 
a little way beyond the end of the vas deferens. The atrium joins the penis, which 
lies in a canal directed forwards and downwards to the male aperture. This canal is 
narrow and cylindrical, wider near its external termination, contracting again, 
however, at the actual aperture, which is situated ventrally on xi, in the position 
of the missing setal sacs. The penis itself is surrounded by a chitinous penis sheath 
lying loosely within the canal. The sheath is tubular, circular in cross section, 
narrowing gradually along its course, but expanding again and curving forwards at 
its outer end ; its length is about 520m, its breadth 49m above, at its narrowest part 
28m. The lower side of its expanded end terminates in a free margin ; the upper side 
is curved strongly upwards and joins the wall of the penis canal (fig. 16) ; in the 
section there illustrated a horizontal cellular shelf projects backwards into the open 
end of the penis sheath. The sheath ends some distance above the external aperture 
of the penis canal. The vesiculse s&minales comprise a pair of sacs in ix, forward 
bulgings of septum f^,, and a long single sac extending back through xi and a number 
of succeeding segments, which is essentially a backward bulging of septum ^. 

The ovaries in segment xi correspond in position to the testes ; they are large, 
extending upwards at the sides of the alimentary canal, and becoming dorsal to it. The 
funnel is situated ventrally on septum y^- ; the oviduct is a minute passage leading to 
the exterior in the inter-segmental constriction between xi and xii. The ovisac extends 
backwards in the same way as the sperm-sac. The spermathecse are in segment x ; 
their external aperture is just in front of and in line with the setal sacs. At first the 
passage runs vertically ; its lumen here is narrow, and its epithelial lining consists of 
deeply staining cubical or low columnar cells, with a cuticle on their free surface. The 
height of the epithelium soon becomes irregular, and the outline of the lumen in conse- 
quence wavy. All this first part, or duct, is invested by a strong muscular covering, 
in two layers, a longitudinal external and circular internal ; its diameter is about 80m. 
After bending about, the cavity widens very considerably ; the lining consists of a 
rather irregular layer of low columnar cells, and the lumen is filled with a granular 
coagulum ; the diameter of the ampulla is about 270m. There were no sperinatophores. 

The diagnosis of the genus Limnodrilus, as given by Michaelsen (5), runs as 
t ; )l lows: — " Ventrale und dorsale Bundel lediglich mit gleichartigen, gabel-spitzigen 



BRANOHIURA S0WERBY1 BEDDARD. 301 

Hakenborsten. Mannliche poren am 11, Samentaschenporen am 10 segm. Supra - 
intestinalgefass und Subintestinalgefass vorhanden ; Transversalgefasse des 8 und 7 
Segm. herzartig ; integumentaler Blutgefassplexus meist vorhanden, aber sparlich. 
Nephridien mit Endblase. Hoden im 10 Segm. ; Samenleiter lang, in das proximale 
Ende der Atrien einmiindend ; Atrien mit einer grossen Prostata ; Penis mit Chitin- 
scheide. Samentaschen im 10 Segm. Spermatophoren in den Samentaschen." 

The present, therefore, differs from other species of the genus in the absence of a 
subintestinal vessel ; in the presence of only one pair of hearts ; in the absence of a 
terminal dilatation of the nephridial canal ; and apparently in the ventral position of 
the dorsal vessel. But having regard to the rest of the anatomy, especially the 
characters of the setae, of the genital organs, including the chitinous penis-sheath, and the 
cutaneous blood-plexns, there can be little doubt that it should be included in the 
genus. 

Of the species within the genus, it resembles most closely L. hojfmeisteri, Clap., 
with which it appears to be very similar in its general appearance, size, number of 
segments, number of setae per bundle, characters of brain and pharynx. None of these, 
however, are particularly distinctive ; more striking are the investment of the anterior 
nephridia with bladder-like (blasenformigen) cells, and the number of cutaneous twigs 
per segment. The general proportions of the penis-sheath are much the same ; but 
the curve of the tube and the character of its mouth differ considerably (cf. Vejdovsky, 
11, pi. xi. fig. 4). The mode of origin of the cutaneous capillaries, the fact that they 
do not end blindly in the present species, and that the dorsal and ventral vessels are, 
as usual, on opposite sides of the alimentary tube in L. hojfmeisteri [cf. Vejdovsky, op. 
cit., pi. viii. figs. 16, 17), on the same side in this species, as well as, presumably, the 
points given above wherein this species differs from all others of the genus, also serve 
to distinguish them. 

I propose the following diagnosis : — Colour pale reddish brown ; length 40-75 mm. ; 
segments about 110, double annulation in the first few; prostomium bluntly conical; 
clitellum xi-^ xii. Setae 6-8 per bundle anteriorly, 3-4 posteriorly. A single pair of 
hearts in viii ; dorsal vessel is ventral in position, to left of ventral vessel, from posterior 
end as far as genital segments ; no subintestinal vessel. Nephridia in vii and viii invested 
with large pyriform peritoneal cells ; no nephridia in ix-xii ; not present in every 
segment from xiii onwards. Cerebral ganglion deeply cleft in front, slightly so behind. 
Chitinous penis-sheath 10-11 times as long as its widest part is broad, curved forwards 
at its lower end, where its anterior lip is strongly reflexed upwards. 



TRANS. ROY. SOC. EDIN., VOL. XLVTII. PART II. (NO. 15). 46 



302 DR J. STEPHENSON ON 



REFERENCES TO LITERATURE. 



(1) Ashworth, J. H., "The Giant Nerve Cells and Fibres of Halla parthenopeia," Phil. Trans. Roy. Soc. 

Land., series B, vol. cc. (1909). 

(2) Beddard, F. E., "A new branchiate Oligochsete (Branchiura sowerbyi)," Quart. Joum. Micr. Science, 

N.S., vol. xxxiii. (1892). 

(3) Beddard., F. E., A Monograph of the order Oligochseta, Oxford, 1895. 

(4) Bourne, A. G., " On Cheetobranchus, a new genus of Oligochaetous Chsetopoda," Quart, Joum. Micr. 

Science, N.S., vol. xxxi. (1890). 

(5) Michaelsen, W., " Oligochseta," in Das Tierreich, Berlin, 1900. 

(6) Michaelsen, "W., "Zur Kenntnis der Tubificiden," Arch, fur Naturgesch., vol. lxxiv., Bd. i., Heft 1, 

Berlin, 1908. 

(7) Michaelsen, W., "Die Oligochatenfauna der vorderindisch-ceylonischen Region," A bh. aus dem Gebiete 

der Natvriv. hgbn. voni Naturw. Verein in Hamburg, Bd. xix., Heft 5 (1910). 

(8) Perrier, Leon, "Une station rhodanienne de Branchiura soiuerbyi Redd.," Ann. de Vuniversite de 

Grenoble, tome xxi., no. i., Paris, 1909. 

(9) Stephenson, J., " Studies on the aquatic Oligochseta of the Punjab, I.," Records of the Indian Museum, 

vol. v., parti. (1910). 

(10) Stephenson, J., "On some littoral Oligochseta of the Clyde," Trans. Roy. Soc. Edin., vol. xlviii., pt. i. 

(1911). 

(11) Vejdovskv, F., System und Morphologie der Oligochgeten, Prag, 1884. 



EXPLANATION OF FIGURES. 

(Figs. 2, 3, 4, 5, 6, 10, 11, 16 drawn by camera lucida.) 

Fig. 1. Single- and double-pointed seta? of Branchiura soiuerbyi. x 450. 

Fig. 2. Section through anterior part of Branchiura soiuerbyi; septum i, which is bulged backwards, 
is cut through in the upper part of the figure. Blood-vessels shaded. x 154. 

circ. m., circular muscular coat; ep., surface epithelium; in. circ. m., inner circular muscular coat; I. I., 
lateral line cells ; long, m., longitudinal muscular coat ; m. set., dorso-ventral muscular bundle between setal 
sacs of same side; ces., oesophagus; p., peritoneal cells ("septal gland"); s., ventral seta? of segment ivj 
sept., septum t, consisting of decussating muscular fibres ; v. n. c, ventral nerve cord, with giant fibre dorsally. 

Fig. 3. Section through gill region of the same ; the anterior face of the section is towards the observer, 
i.e. left and right are reversed. The longitudinal muscular layer is here broken up, leaving visible a fibrillar 
groundwork with a few nuclei. x 250. 

c, a group of cells connected with the ventral vessel, prominent in this particular section ; chl., chlora- 
gogen cells ; d. v., dorsal vessel, showing a connection with the enteric plexus ; g. /., giant fibre, to the left 
of which, in the figure, are seen a row of similar but smaller fibres; int., intestine; plex., part of intestinal 
blood-plexus ; v. v., ventral vessel. Other references as above. 

Fig. 4. Cells of the lateral line in the posterior part of the same, in transverse section. The lateral 
line is here discontinuous ; where present the cells project some distance inwards and spread out. x 560. 

cut., cuticle; int. ep., intestinal epithelium ; int. m., intestinal muscular coat; per., peritoneal cells round 
intestine. Other references as above. 

Fig. 5. Transverse section of the same near the extreme posterior end, showing the transverse muscular 
partition, the absence of a definite body-cavity in the ventral part of the section, and the relatively large size 
(if the ventral nerve cord. x 250. 

m., bands of muscle fibres, in various directions; transv., transverse band (the coelomic partition); x., 
groundwork of fibrillse and nuclei. Other references as before. 



BRANCHIURA SOWERBYI BEDDARD. 303 

Fig. 6. Section along a ventral gill, to show epithelial covering, continuation of circular muscular layer 
of body on the gill, gill cavity, and cells contained within it. x 250. 

a., branching cells within gill cavity : b. v., blood-vessel ; cav., gill cavity ; m., muscular strands coming 
down from the transverse partition. Other references as before. 

Fig. 7. Branchiura sowerbyi ; blood-vessels to the gills; from a sketch from life, from the right side ; 
the dorsal vessel is therefore behind the intestine as seen by the observer. Of the two cutaneous branches 
shown, the one on the right is coming towards the observer and is superficial, the one on the left is on the 
deep surface. 

cut., cutaneous branch; d. g., dorsal gill; d. v., dorsal vessel; int., intestine; v. g., ventral gill; v. v., 
ventral vessel. 

Fig. 8. Transverse section through gill region of Branchiura sowerbyi ; diagrammatic, constructed from 
sections to show the course of a number of vascular branches, which are not on the same level. The dorsal 
vessel d. v. gives off a dorsally directed branch d. d. to the dorsal gill, and a ventrally directed branch v. d. 
to the ventral gill. The ventral vessel v. v. gives off a branch on the right side which bifurcates into a dorsal, 
d. br. v., and a ventral division, v. br. v., going to the dorsal and ventral gills respectively. The ventral 
vessel also gives off a branch on the left side, I. br. v. 

d. g., dorsal gill; ep., surface epithelium; int., intestine; i. p., intestinal plexus; I. to., longitudinal 
muscular layer (the circular layer is not separately represented) ; v. g., ventral gill ; v. n. c, ventral nerve 
cord. 

The section is seen from its anterior face ; right and left are therefore reversed. 

Fig. 9. Seta of Limnodrilus socialis. x 750. 

Fig. 10. Part of transverse section through the fourth segment of Limnodrilus socialis, showing the 
lateral portion of the body-wall. x 250. 

circ. m., circular muscular layer ; cut., cuticle ; ep., surface epithelium ; I. I., lateral line cells ; long, to., 
longitudinal muscular layer; to. set., muscular bundle between setal sacs of the same side ; per., peritoneal 
cells ; s., setae. 

Fig. 11. Transverse section through sixth segment of the same. x 115. 

chl., chloragogen cells; d. v., dorsal vessel; gl., gland-like masses of cells; oes., oesophagus with ventral 
ridge projecting into lumen ; s. i. v., supra-intestinal vessel ; v. n. c, ventral nerve cord ; v. v., ventral vessel. 
Other references as above. 

Fig. 12. Portion of vascular system of the same, from the right side; from a sketch from the living. 
The anterior and posterior parts of the ventral vessel are continuous hy means of a fine channel ; the trans- 
verse sinus is seen in ix, and a part of the contractile loop to the genital organs. 

Fig. 13. Portion of vascular system of the same from above, showing the relations of dorsal and supra- 
intestinal vessels, the hearts, the transverse sinus and its associated plexus ; from a sketch from the living. 

Fig. 14. Blood-vessels on the left side of the intestine of the same; from a sketch from the living; the 
anterior end is towards the left. The dorsal vessel is shown below the intestine. 

Fig. 15. Outline of cerebral ganglion of the same. 

Fig. 16. Longitudinal section through lower part of penis tube and sheath of the same, to show the 
shape of the lower end of the chitinous sheath and its relations. x 250. 

circ. m., circular muscular layer; ep., surface epithelium; long, to., longitudinal muscular layer; per., 
peritoneal cells ; s., penis sheath ; t., penis tube ; v. def., vas deferens ; x., shelf-like projection into lower end 
of penis sheath. 



Traii. Roy. Soc. Edin' Vol. XLVIII. 

Stephenson : Branchiura sowerbyi, Beddard. -Plate T. 

eire.in.. 

urt.; 

a ■■[■■ 




J:f tt n ; c - ^ ' &Hfl. »l . 

/ / Arc. >n- 



M?F*rlone & Ersklne. Edu 



^svTwt'i 



^ 



t"?AL H\ 



H Koy- Soc - Edin ' Vol XLVIII. 

Stephenson: Branchiura sowerbyi, Beddard.-Plate II. 







16 




14 




M'Firlane & Ersklne. Edn 



( 305 ) 



XVI. — The Tunicata of the Scottish National Antarctic Expedition, 1902-1904. 
By W. A. Herdman, D.Sc, F.R.S., Professor of Zoology in the University of 
Liverpool. (With One Plate.) 

(MS. received January 8, 1912. Read February 19, 1912. Issued separately July 3, 1912.) 

So far as regards number of individuals, and their size, this is one of the largest 
collections of Ascidians brought back in recent years from Antarctic seas. It contains 
almost exactly the same number of species of Ascidiacea (Ascidiae Simplices + Ascidise 
Composite) as the Discovery collection — viz. fifteen or sixteen — but whereas in the 
latter collection nearly all the species were represented by single specimens, in the 
Scotia collection most species can show long series of individuals — in all there are about 
two hundred specimens, as against the thirty-three brought home by the Discovery. 

The sixteen species in the present collection represent almost as many genera, and 
half a dozen families. The systematic arrangement is as follows : — 



ASCIDIACEA. 

Family Molgulid^e. 

Paramolgula gregaria (Lesson). 
Paramolgula horrida (Herdman). 

Family Cynthiid^e. 

Boltenia legurnen, Lesson. 
Fungulus antarcticus, n. sp. 
Halocynthia setosa, Sluiter. 

Family Styelid^e. 

Styela lactea, Herdman. 
Styela paessleri, Michaelsen. 
Synstyela incrustans, Herdman. 



Polyzoa opuntia, Lesson. 
Goodsiria placenta, Herdman. 

Family Ascidiid^e. 
Ascidia charcoti, Sluiter. 

Family Distomid,e. 

Colella pedunculata (Q. and G.). 
Holozoa cylindrica, Lesson. 

Family Polyclinid^. 

Polyclinum complanatum, Herdman. 
Amaroucium distomoides, Herdman. 
Amaroucium sp. 



It is interesting to notice how greatly some of these recent collections from the far 
South differ from one another in the species represented. The following table — which 
gives only the sixteen species in the Scotia collection — shows that only one form 
(Halocynthia setosa) from that collection was also taken by the Discovery, whereas ten 
species were taken by the Challenger, eight by the Hamburg Magellanic and South 
Georgia Expedition, and five by the French Antarctic Expedition under Charcot. This 
can be explained to some extent, at least, by the precise localities visited : the Scotia, the 
Cliallenger, and the Hamburg collections were largely made in the Magellan and Falk- 
lands neighbourhood, while the other three collections were mainly from farther south. 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 16). 47 



306 



PROFESSOR W. A. HERDMAN ON THE 



" Scotia " Species. 


"Discovery." 


"Challenger." 


Charcot. 


"Southern 
Cross." 


Magellan 

and South 

Georgia. 


Paramolgula horrid a, 




X 






X 


„ gregaria, 




X 






X 


Boltenia legumen, . 




X 






X 


Fungulus antarcticus, 


("Scotia" 


only) 








Halocynthia setosa, . 


X 




X 






Styela lactea, . 




X 


x 


x 




„ paessleri, 










X 


Synstyela incrustans 




X 






X 


Polyzoa opuntia, 




X 






X 


Goodsiria placenta, . 




X 








Ascidia charcoti, 






X 






Colella pedunculata, 




X 


X 




X 


Holozoa cylindrica, . 




X 


X 


X 


X 


Polyelinum complanatum, 


) 








A maroucium distomoides, 


> Iu collection of Australian Museum, Sydney. 






Amaroucium sp., 


) 









Although only one of the Scotia Tunicata requires to be described as new to science, 
several of the species are of considerable interest, and most of them add something to 
our knowledge either in the characters and variation of the species or in distribution. 
The one new species [Fungulus antarcticus) is a very remarkable form belonging to the 
deep-sea genus Fungulus, known only from a single species obtained during the 
Challenger Expedition between the Cape of Good Hope and Kerguelen Island. 

This collection shows again what I remarked upon more than twenty years ago in 
the case of the Challenger collection, that the Ascidian fauna of the far South is 
characterised by the abundance and the large size of the individuals of a comparatively 
few species. Halocynthia setosa and Holozoa cylindrica are the two largest species, the 
one simple and the other compound, and both are represented by a large number of 
specimens. I have, however, written on this matter, and also on the number of Antarctic 
as compared with Arctic species, so recently in my report * upon the Discovery collection 
that these matters need not be discussed further here. 



Family Molgulid^e. 

Paramolgula gregaria (Lesson). (Plate, fig. 9.) 

Cynthia gregaria, Lesson, Cent. Zool., p. 157. 

Molgula gregaria, Herdman, Challenger Report on Tunicata, Part I., p. 73. 

Locality. — Station 118, on hulks, Stanley Harbour, Falkland Islands, January 
16, 1903. 

There are over forty specimens of this species in the collection, ranging in size from 
2x T5 cm. up to 6'5x5 cm. The majority are about 4 cm. in diameter. They have 

* National Antarctic Expedition: Natural History, vol. v., "Tunicata," 1910. 



TUNICATA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 307 

the characteristic smooth test and translucent grey tint, and they also agree closely 
in internal details with the Challenger specimens from shallow water off the Falkland 
Islands. 

Michaelskn has suggested that this species and Paramolgula gigantea (Cunningham) 
are the same. No doubt they are related forms ; both belong to the restricted genus 
Paramolgula, having only broad, ribbon-like longitudinal bars but no true folds in the 
branchial sac (a fact I overlooked in drawing up my " Eevised Classification of the 
Tunicata" in 1891, as Michaelsen has pointed out), but I do not consider them as 
identical. In addition to differences in the external appearance — the shape and the 
condition of the test — the branchial sacs are not alike in detail, and the dorsal tubercles 
differ widely. I give here a figure of the dorsal tubercle (Plate, fig. 9) of P. gregaria 
from the Scotia collection for comparison with that of P. gigantea figured in the 
Challenger Report. 

Lesson figures his species with five lobes round each aperture, but that is no doubt 
an error. The branchial aperture has six and the atrial four lobes. 

Paramolgula horrida (Herdman) (?). (Plate, figs. 10 and 11.) 

Locality. — Station 118, on hulks, Stanley Harbour, Falkland Islands. 

I have very little doubt that this single large specimen (measuring 7*5 cm. in length 
and 5 '5 cm. in breadth) belongs to the same species as the specimen from "off the 
Falkland Islands, 5-12 fathoms," which I described in the Challenger Eeport as Molgula 
horrida. They both fall within the more modern genus Paramolgula, separated off 
from Molgula by Traustedt because of the absence of true folds in the branchial sac. 
As the Challenger description was drawn from a single specimen, and as this Scotia 
specimen differs a little in detail, it may be well, in the interests of a fuller knowledge 
of the species, to add a few of the characteristics of the individual before me. 

The shape is irregularly ovate, and flattened, and the colour is a very dark 
brown. The other external characters can be seen from the figure (Plate, fig. 10). 
The Test is leathery and rough on the surface. It is thin but tough, and dark 
but smooth and glistening on the inner surface. The Mantle is dark brown and 
opaque. It is thick, but soft and not muscular, or at least the muscles do not form 
obvious bands. 

The Branchial Sac has on each side seven wide longitudinal vessels which look like 
narrow folds. Between the distant wider transverse vessels, narrower intermediate ones 
branch in all directions in a dendritic manner, so as to form rounded and oval and 
variously shaped meshes in which the stigmata lie. The stigmata are also rather 
irregular in arrangement, being in some parts in spirals and in other places side by side 
in rows (see fig. 1 1 ). 

The Tentacles are of different sizes, there being eight larger much branched, with 
some smaller ones between. 



308 PROFESSOR W. A. HERDMAN ON THE 

The Dorsal Tubercle has its long axis antero-posterior and its opening at the side. 
The horns form two close spirals, both coiled inwards. Whether P. patagonica of 
Michaelsen is also the same species as P. horrida is very doubtful. I am inclined to 
regard it as distinct. 



"ST 



Family CYNTHiiDiE. 
Boltenia legume?i, Lesson. 

Locality. — Station 118, on hulks, Stanley Harbour, Falkland Islands, about twenty 
specimens ; and eight specimens from 6 fathoms, Port Stanley, February 2, 1904. 

There are over two dozen specimens of this species in the collection, and they range 
in size from 1*5 to 7 cm. in greatest length of body — about the same range as in the 
case of those in the Challenger collection. All of these specimens of B. legumen belong 
to the " forma typica " of Michaelsen's system * of subdivision of this species, and 
agree in character with the Challenger specimens from the same locality. In some 
cases the little bristles on the surface of the test are more abundant and more prominent 
than in others, but there are all gradations between. This is evidently a very common 
Ascidian in shallow water at the Falklands. 



Fungulus antarcticus, n. sp. (Plate, figs. 15 to 18.) 

A single specimen which clearly belongs to the rare and interesting genus Fungulus 
was obtained at Station 301 from a depth of 2485 fathoms, on March 13, 1903, at 
lat. 64° 48' S., long. 44° 26' W. ; temp. 31° "02. The genus was established in 1882 for 
another solitary individual found in the Southern Ocean during the Challenger Expedi- 
tion, at Station 1 47, between the Cape of Good Hope and Kerguelen Island, lat. 46° 
46' S., long. 45° 31' E. ; depth, 1600 fathoms, on a bottom of Globigerina ooze. The 
two localities are thus nearly 3000 miles apart, but agree in that both are in the far 
south and in very deep water. 

The Scotia, specimen, although closely related to the Challenger Fungulus cinereus, 
Herdman, cannot be placed in the same species. The general appearance and anatomy, 
and especially the remarkable structure of the branchial sac, are the same ; but the 
relation of the peduncle to the apertures and the details of structure are different in 
the two forms. The description of F. antarcticus is as follows : — 

The body is club-shaped (fig. 15), like a rounded knob about T5 cm. in diameter 
on the summit of a short, stout peduncle, which is also about 1*5 cm. in length and 
from 4 to 6 mm. in thickness. The peduncle is continuous with the ventral edge of 
the body, while the dorsal edge projects markedly. The surface is smooth and the 

* " Die Holosomen Ascidien des magalhaeniseh-sudgeorgischen Gebietes," Zoologica, Bd. xii., Heft 31, Stuttgart, 
1900, p. 109. 



TUNICATA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 309 

colour pale yellowish grey. The branchial aperture is a little way along the ventral 
edge of the anterior end, and appears to be bilabiate or elliptical rather than lobed. 
The atrial aperture is in the middle of the dorsal surface, and is a square or four-lobed 
opening. In F. cinereus, the branchial aperture was described as triangular, and the 
atrial as bilabiate, but the figure of the former in the Challenger Report is not very 
different from the figure now given (fig. 16) from the Scotia specimen. 

The Test is thin and membranous, but tough. Under the microscope it is seen to be 
finely roughened all over the outer surface. In minute structure the test agrees with 
that of F. cinereus as described in the Challenger Report. The Mantle is very thin and 
inconspicuous, but muscular. It is penetrated by numerous, very fine, closely placed 
muscle bundles which, in the tubular extension of the mantle which occupies the hollow 
peduncle, run longitudinally parallel to one another. 

The Branchial Sac is remarkably delicate, and is, in fact, merely a very loose wide- 
meshed net with folds at intervals where the longitudinally-running vessels are crowded 
together (fig. 17). The transverse vessels are of two sizes, occurring alternately. The 
looseness of the branchial sac and the minute undulations in practically all the muscle 
bundles of the mantle give the impression that when alive and filled with sea- water the 
animal had the power of expanding to a considerably larger size than it now shows. 
Possibly the test when alive was of a gelatinous consistency and capable of being- 
dilated. 

There are no spicules in the vessels of the branchial sac. The endostyle is narrow 
but conspicuous ; there are no spicules in its wall. The branchial tentacles are few and 
only slightly branched. The alimentary canal is relatively small, and is confined to 
the posterior end of the left side close to the top of the peduncle (fig. 18). The stomach 
wall has a number of close-set longitudinal folds. 

The gonads are one on each side, rather long and irregular, with the narrower end 
pointing to the atrial aperture (fig. 18). 

This new species differs from Fungulus cinereus, Herdman, in the shape and pro- 
portions of the body (see figures) and in the much paler colour of the test ; in the 
details of position and shape of the branchial and atrial apertures ; in having the trans- 
verse vessels of the branchial sac distinctly of two sizes ; in having a well-marked 
stomach with longitudinal folds ; and in the length and shape of the gonads. 

Halocynthia setosa, Sluiter. 

This very striking and characteristic species was obtained by the Scotia in consider- 
able quantity at the South Orkneys. It was originally described by Sluiter * from two 
specimens obtained by the French Antarctic Expedition under Dr Jean Charcot at " He 
Booth Wandel, 40 metres " ; but as the figures in the report on the Charcot Expedition 
did not seem to me to be characteristic, I gave a supplementary description, with figures, 

* Bull. Mm. Hist. Nat. Paris, 1905, No. 6, p. 472 ; and Expe'd. Antarct. Franc. (Charcot), " Tuniciers," p. 40. 



310 PROFESSOR W. A. HERDMAN ON THE 

of the species in my report upon the Tunicata of the Discovery Expedition.'"" The 
Discovery obtained in all five specimens from the east end of the Barrier and the 
neighbourhood of the winter quarters in M c Murdo Bay, in 10-20 fathoms. 

The more abundant material obtained by the Scotia gives us still further informa- 
tion in regard to the characteristics and variation of the species. Some of the Scotia 
specimens, measuring up to 11 x 7 x 5 cm., are the largest yet obtained; and some of 
them show a short peduncle at the place of attachment, a feature not previously 
observed. The characteristic spines on the test reach in some of these larger specimens 
to a length of 21 mm. and a breadth of about 1 mm. at the base. In some of these 
specimens the musculature of the mantle is remarkably strong, and consists externally of 
circular siphonal sphincters, beyond which is an oval sphincter, of numerous fibres, 
enclosing both siphons, while more internally radial muscles formed of exceedingly 
stout and strong fibres run outwards from the base of the siphons. Connective tissue 
permeated by fine fibres covers and unites all these various muscle bundles. 

Of the six folds on each side of the branchial sac, the largest is the one next to the 
dorsal lamina on each side, and the smallest is generally the third counting from the 
dorsal to the ventral edge. The transverse vessels are of four different sizes, and the 
stigmata are from nine to twelve in a mesh. One dorsal tubercle examined measured 
4*8 mm. from side to side and 3 "2 mm. antero-posteriorly. 

The nerve ganglion is extraordinarily narrow and elongated, and may reach 9*5 mm. 
in length, with two nerves diverging from each end which can be traced with the eye 
round the sphincters of both siphons. The subneural gland is in the form of a thin 
layer over the ganglion. 

A strong band of muscle fibres lies under the dorsal lamina and extends from the 
mantle into the wall of the branchial sac near the posterior end of the nerve ganglion. 

Two of the specimens had Amphipods in the branchial sac. The specimens were 
obtained as follows : — 

I. Station 325,t 9-10 fathoms, Scotia Bay, South Orkneys, July 1903. 

( 1 ) 9x6x6 cm. (on a short peduncle). 

(2) 9'5 x 6*5 x 3 cm. (very short peduncle). 

(3) 8x5-5x4 cm. 

11. Station 325, 9-10 fathoms, June 1903 ; temp. 29° F. 

(1) 4x4 "5x2 - 5 cm. (test only). 

(2) 8 x 5*5 x 4 cm. (with a smaller one attached). 

III. Station 325, 9-10 fathoms, August 1903 ; temp. 29° F. 

(1) 6 x 5 x 4'5 cm. (also four empty tests). 

IV. Station 325, 9-10 fathoms, May 1903; temp, about 28° F. 

(1) 7*4x5x5 cm. 

(2) 6x4x4 cm. (also four empty tests). 

* Report National Antarctic Exped. : Nat. Hist., vol. v., "Tunicata," London, 1910. 
t The whole of Scotia Bay is termed Station 325 ; consequently, depths vary. 



TUNICATA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 311 

V. Station 325, 9-15 fathoms, April 1903 ; temp. 28° to 29° F. 

About a dozeri specimens ranging from 11x7x5 cm. down to 5*5 x 5 x 3*2 
cm. (one empty test). 

VI. Station 325, 9-15 fathoms, December 26, 1903 ; temp. 3L°'4 F. 

(1) 4^x3x3 cm. 

Family Styelid^e. 

Styela lactea, Herdman. (Plate, figs. 1-8.) 

Styela lactea, Herdman, Challenger Keport on Tunicata, Part I., p. 156. 
Styela flexibilis, Sluiter, Charcot Exped., "Tuniciers," p. 36. 
(?) Cynthia verrucosa, Lesson, Cent. ZooL, p. 151. 

Localities. — (1) Station 118, on hulks, Stanley Harbour, Falkland Islands. 

(2) Scotia Bay, South Orkneys, Station 325, February 2, 1904. 

(3) Attached to Holozoa cylindrica, thrown up on beach, Scotia Bay, January 
17, 1904. 

(4) Station 118, shore pools, Port Stanley, January 1903. (Two elongated 
specimens.) 

The specimens from the Falkland Islands are about twenty in number, ranging from 
little globular spiky balls (see figs. 3, 4) of 1 cm. in diameter to irregular barrel-shaped 
masses (fig. 1 ) of 8 cm. in length and 5 to 6 cm. in breadth. The specimens from 
Scotia Bay attached to the compound Ascidian Holozoa cylindrica, Lesson ( = Distaplia 
ignota, Herdman), are small and globular, bristling with short pointed spikes, and of a 
pure white colour (fig. 2) ; while the remaining specimen from Scotia Bay (February 2) 
is much larger, roughly cylindrical in shape, less spiny, and of a duller colour (fig. l). 
Still, all transitions in shape and appearance can be found between the extreme forms, 
so there can be no doubt that all belong to tjie one species, S. lactea, found by the 
Challenger Expedition at Kerguelen Island, and by the Southern Cross Antarctic 
Expedition at Cape Adare. 

The largest Scotia specimens correspond closely with Sluiter's S. flexibilis, found 
during the Charcot Expedition at " He Booth Wandel." That species agrees in internal 
characters with S. lactea so closely that I have no doubt that the two are the same, 
and that S. flexibilis must be regarded as a synonym of S. lactea. It is, I think, 
possible also that the Cynthia verrucosa of Lesson, found attached to Fucus on the 
shores of Malonines Islands, Antarctic, which is figured as having five lobes round each 
aperture, is really this same species. If so, the number of lobes shown by Lesson is, of 
course, erroneous. 

The following additional characters, taken from the larger Scotia specimens, may be 
useful to compare with the descriptions of other specimens : — 

Size 7 x 4 x 3 "5 cm. Barrel-shaped, attached by flat area at posterior end about 
3 '5 cm. in diameter. Colour pale creamy white with a pinkish tinge in places. Test 



312 PROFESSOR W. A. HERDMAN ON THE 

thin, leathery, raised at intervals to form little pointed tubercles, the larger of which 
are echinated (fig. 5). Mantle muscular, with regular circular and longitudinal bands. 
Branchial Sac with four large folds on each side. There are six to nine bars on a fold, 
and four in the interspace. Dorsal lamina a broad plain membrane. There are about 
thirty very long simple tentacles and some intermediate smaller ones. The dorsal 
tubercle has both horns coiled inwards to form short spirals (fig. 8). There are two or 
three long gonads on each side, and many endocarps. Fig. 7 shows the arrangement 
of the alimentary canal. 

In the smaller, more globular specimens the conical spiny tubercles on the test are 
relatively more numerous and more closely and regularly placed (see figs. 2, 3, 4, and 6). 
The Challenger specimen figured, from Kerguelen, was intermediate in size to the 
larger and the smaller Scotia examples, and was smoother in character of test. 

Styela paessleri, Michaelsen. (Plate, figs. 12 to 14.) 

This species was described by Michaelsen in 1900 from specimens obtained in the 
Straits of Magellan. The Scotia specimens from the Falkland Islands seem to be 
rather larger on the whole, but agree in essential characters. 

The following description, from the Scotia material, may be useful : — There are 
about twenty specimens, varying in size from 1 cm. to 3 cm. in length by 1*5 cm. in 
average breadth, obtained from Station 118, at the Falklands, depth 6 fathoms; and a 
couple from Port Stanley, February 2, 1904, 6 fathoms. 

The colour varies from a creamy white to a yellowish brown, and the surface of the 
test is in most places closely wrinkled. The branchial sac has four folds on each side, 
the largest being those adjacent to the dorsal lamina, with ten bars each, while the others 
have usually six bars. Fig. 12 gives the plan of both sides of the branchial sac as seen 
in section, with the number of bars and of rows of stigmata shown by the figures. The 
folds have from five to ten bars, and there are from two to five (usually four) bars in the 
spaces between. These numbers agree fairly well with those given by Michaelsen. 
The transverse vessels are of three sizes arranged with regularity : 1 — 3 — 2 — 3 — 1, and 
having a narrow horizontal membrane in addition crossing the meshes (fig. 13). Most 
of the meshes are square, with five to seven stigmata in each. The extreme dorsal and 
ventral meshes are more elongated transversely, and contain a greater number of stigmata. 

The dorsal tubercle is of curious form (fig. 14), a simple crescent with the horns 
anterior and having a globular excrescence in the concavity. The dorsal lamina is a 
plain membrane. The tentacles are crowded and number about a hundred. They are 
of two sizes, roughly fifty of each. Michaelsen records only sixty tentacles, but as 
the specimens he examined were smaller than ours, the difference need not be regarded 
as important. 

Although some of the above characters do not agree precisely with those given by 
Michaelskn, still the differences are not, I think, greater than what may reasonably he 



TUNICATA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 313 

ascribed to individual variation within the limits of a species. The dorsal tubercle is 
perhaps the feature that shows most divergence, but Michaelsen himself remarks in 
the original description that it is probable that other specimens might show a different 
form of tubercle. 

Polyzoa opuntia, Lesson, subspecies coccinea, Cunningham. 
Goodsiria coccinea, Cunningham, Trans. Linn. Soc. Lond., xxvii. 
Several specimens of this common species were obtained at the Falklands : — 

(1) Station 349, shore pools, Cape Pembroke, January 1903 to January 1904. One 

large, lobed colony and a couple of small ones. This is part of collection 
made on behalf of Scotia by Mr Pearson, Cape Pembroke lighthouse-keeper, 
during twelve months. 

(2) Station 118, rock cod trap, Stanley Harbour, 3^ fathoms, January 1903. One 

elongated colony, about 26 cm. in length. 

Goodsiria (Gynandrocarpa) placenta, Herdman. 

Several specimens that seem to agree closely with this South African species were 
obtained at the Falklands, as follows : — 

(1) Station 118, Stanley Harbour, January 7, 1903. One small colony. 

(2) Station 118, rock cod trap, Stanley Harbour, 3^ fathoms, January 1903. Part 

of a large colony which probably measured 10 or 12 cm. across. 

(3) Station 118, Port Stanley, 6 fathoms, February 1904. One colony measuring 

about 10 by 5 cm. 

Synstyela incrustans, Herdman. 
(1) Allceocarpa zschaui (Michaelsen). 

Locality. — Station 118, on hulks, Stanley Harbour, Falkland Islands. 

There are about a dozen colonies of this species, ranging in size from 1 or 2 cm. up 
to 5 or 6 cm. in diameter. Most of them were adhering in masses along with the larger 
specimens of Paramolgula gregaria. 

In detailed characters these specimens agree well with the Challenger specimens of 
Synstyela incrustans obtained in the Straits of Magellan, but they also agree with 
Michaelsen's description of Allceocarpa zschaui from South Georgia ; and when 
mature, the Ascidiozooids show the male unisexual polycarps on the left, and the female 
on the right-hand side of the mantle, which is a character of Michaelsen's proposed 
generic division Allceocarpa. As, however, he names my species Synstyela incrustans 
as the type form of Allceocarpa, and as he apparently does not in his system retain 
Synstyela as a genus, but substitutes the name Allceocarpa for it, I must point out 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 16). 48 



314 PROFESSOR W. A. HERDMAN ON THE 

that Synstyela, Giard, has the prior claim and must be retained as the name of the 
genus, even when, as happens to be the case, our knowledge of the internal characters 
has been increased and the definition added to since the genus was originally created. 
Consequently I must regard Michaelsen's Allceocarpa zschaui as a Synstyela, and 
furthermore I find myself unable to distinguish it as a species from S. incrustans of 
the Challenger Report. In Michaelsen's "Revision der compositen Styeliden oder 
Polyzoinen," # where both species are described, in his table on p. 73 he distinguishes 
them by the proportions of the oviduct and the number of internal longitudinal bars in 
the branchial sac, as follows : — 

S. zschaui having the oviduct broader than long, and having sixteen to seventeen 

bars on each side ; and 
S. incrustans having the oviduct longer than broad, and having twelve to fourteen 

bars on each side of the sac. 

Now, in the first place, with a soft, easily deformed structure like the oviduct it is 
almost impossible to be sure of the true proportions ; and secondly, I find them varying 
considerably in my specimens ; so that I cannot say they agree more in this character 
with the one species than with the other. Then as to the number of longitudinal bars, 
on dissecting out and mounting a branchial sac from a Scotia specimen I find the 
number of bars to be fifteen on each side. According to Michaelsen, if it had sixteen 
the species would be zschaui, and if it had fourteen it would be incrustans. Under 
these circumstances, and as I find the specimens before me agree equally well with the 
descriptions of these two species, 1 think there can be little doubt but that A. zschaui, 
Michaelsen, is a synonym of Synstyela incrustans, Herdman. 

Diandrocavpa monocarpa (Sluiter) is certainly not the same species as Synstyela 
incrustans, although it is probably a Synstyela. The number of longitudinal bars in 
the branchial sac is very much smaller than in the present species. 

Family Ascidiid^e. 

Ascidia charcoti, Sluiter. 

Locality.— Station 325, in shore pool, Scotia Bay, South Orkneys, February 2, 
1904. 

The single large Ascidia in the collection clearly belongs to Sluiter's A. charcoti, 
a species found by the Charcot Expedition to be abundant at " He Booth Wandel." 
The Scotia specimen measures 8*5 x 5*5 x 2 cm., and was attached by a small area in 
the middle of the left side. The branchial aperture has only seven lobes, a curious little 
detail in which it agrees with Sluiter's description. The atrial has the usual six lobes 
characteristic of the genus. The test reaches a thickness of 2 to 3 mm., but has not 
the red colour mentioned by Sluiter ; and the mantle is unusually thick and spongy 

* Mitleilungen aus dem Naturhistor. Museum, xxi., Hamburg, 1904. 



TUNICATA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 315 

for an Ascidia. The branchial sac is also thick, and both mantle and branchial sac are 
of a distinctly pinkish colour which may be the remains of the orange-red that Sluiter 
records. There are twelve moderate-sized tentacles, and the horse-shoe shaped dorsal 
tubercle is very large, reaching up to the base of the tentacles. It seems larger than in 
Sluiter's specimens, in which, however, the dorsal tubercle is recorded as being rather 
variable. 

Sluiter states that no teeth are present on the dorsal lamina ; but I find that in 
the Scotia specimen the dorsal lamina has marked denticulations along its free edge, 
amounting in one part to short tentacular languets. But still I have no doubt that 
my specimen belongs to Sluiter's species, and that the dorsal lamina must be regarded 
as somewhat variable in character. The viscera on the left side of the body are 
unusually large and opaque. 

Family Distomid^e. 

Colella pedunculated (Quoy and Gaimard). 

1 Sycozoa siyillinoides, Lesson. 
? Colella tenuicaulis, Herdman. 
? Colella umbellata, Michaelsen. 

One colony having a stalk bifurcated near the top and bearing two " heads " was 
found at Station 346 on Burdwood Bank, 56 fathoms, on December 1, 1903, and presents 
to some extent characters recalling all the species named above. In the branching of 
the peduncle it is like Michaelsen's C. umbellata from the Falklands ; in the general 
appearance of the " head," however, it is more like Quoy and Gaimard's C. pedunculata, 
found by the Challenger at the Straits of Magellan, the Falkland Islands. Kerguelen, 
etc. The long slender stalk recalls the Australian C. tenuicaulis ; and it is possible 
that Lesson's Sycozoa sigillinoides may be identical with one or more of these other 
named forms. Both the "heads" are, unfortunately, in the single colony in a very 
ragged condition — possibly dead when collected — so that the more minute characters 
of the Ascidiozooids cannot be determined. 

Holozoa cylindrica, Lesson. (Plate, fig 2.) 
-(?) ignotus, Herdman, Challenger Report, ii., 1886, p. 251. 



Julinia australis, Caiman, Quart. J. Micr. Set., 1894, p. 1. 

Distaplia ignota, Herdman, Report on "Southern Cross" Tunicata, Brit. Mus., 1902, p. 197. 

Holozoa cylindrica, Less. — Hartmeyer, in Bronn's Tier-Reichs, "Tunicata," 1909. 

This large and apparently abundant Antarctic species was obtained by the Scotia 
Expedition at the following localities : — 

I. Station 346, Burdwood Bank, 56 fathoms, December 1, 1903. Seventeen 
fragments from 10 to 30 cm. in length by 2 to 4 cm. in diameter. All 
in bad condition, soft and partly macerated, with many other animals, 
Hydroids, Polyzoa, etc., entangled in the irregular, ragged surface. 



316 PROFESSOR W. A. HERDMAN ON THE 

II. Station 325, Scotia Bay, April 1903. One specimen, 30x4x3 cm., bad 
condition. 

III. Station 325, Scotia Bay, South Orkneys, December 6, 1903; temp. 29"8° ; 

floating on surface. 80 cm. x 2 (tapering to 1) cm. 

IV. Station 325, Scotia Bay, South Orkneys, December 26, 1903 ; temp. 307°. 

(1) 85 (incomplete) x 1'5 (tapering to 1) cm. 

(2) 75 cm. (incomplete) and two fragments. 

V. Station 326a, Brown's Bay, South Orkneys, November 1903. Two specimens : 
(1) 55 x 2 to 3 cm ; (2) 40 x 2 cm. 
VI. Scotia Bay, South Orkneys, January 17, 1904 ; temp. 3 2 "5° ; thrown up on 
beach. One colony, 20 x 5 cm., with several Styela lactea attached ; in 
bad condition ; most of Ascidiozooids lost. 
VII. Scotia Bay, South Orkneys, January 3, 1904; temp. 31*5°; thrown up on 
beach. Two very long specimens : (1) over 100 x 2 cm. ; (2) over 
150 x 2 cm. 
VIII. Scotia Bay, South Orkneys, November 25, 1903; surface. Three small 
colonies, 20 to 30 x 1 to 1*5 cm. 
IX. Scotia Bay, March 25, 1903. One small colony, 10 x 2 cm. ; bad condition; 
most of Ascidiozooids gone. 
Most of these specimens are, unfortunately, in very bad condition, and were probably 
dead or decomposing when collected. The Challenger specimens were in such a rotten 
condition that it was impossible to determine even the genus. But from the rather 
better material brought home by the Southern Cross Expedition I was able to determine 
that the Challenger specimens — evidently the same species — belonged to the genus 
Distaplia. What Calman described as Julinia australis in 1894 is again the same. 

Sluiter, in his report on the Charcot Tunicata, thinks that "Julinia" may be 
recognised as an independent genus because of the elongated form of the colony ; but 
Distaplia clavata (Sars), from Arctic seas, although it does not attain to such a length, 
has the same elongated form — and therefore it cannot be said that a Distaplia with this 
habit of growth is unknown. 

The colony found floating on the surface in Scotia Bay, December 26, and measuring 
about 85 cm. in length, is the best preserved specimen in the collection, and I think 
the best preserved that I have seen in any collection brought back from the Antarctic. 
The colony, although soft, does not seem to be rotten. The Ascidiozooids are distinct 
and large and closely placed throughout its length. Their exposed ends measure about 
2 mm. across, and are of an opaque pale yellow colour, in contrast to the translucent grey 
of the test in which they are embedded. Throughout the greater part of the colon)' the 
Ascidiozooids appear to be in long meandering lines, but here and there one comes upon 
a circular, elliptical, or more irregular group (fig. 2), reminding one of the arrangement 
in a Botrylloides. Both ends of the colony are incomplete, and at the upper end the 
Ascidiozooids appear to be dropping out of the test. 



TUNICATA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 317 

Some of the Ascidiozooids in this colony are the best preserved I have seen in all the 
various samples of this species that have passed through my hands, and their anatomical 
and histological characters agree in detail with the excellent account of " Julinia " given 
by Calman. In fact, I can agree with Calman in every respect save that of bestowing 
a new name on the genus. It is evident from his remarks that he recognised the close 
affinity to Distaplia, and the only mistake he made was in not referring the species to 
that genus. 

I agree, however, with Hartmeyer* that it is practically certain that this form had a 
distinctive generic name applied to it at a still earlier date. The " Holozoa cylindrica " 
of Lesson (Voyage " Coquille," Zool., ii. p. 439 ; 1830) agrees in all the points that are 
mentioned in the brief description with our form. It is said to have a " holothuriform " 
body, cylindrical, with rounded ends, free and floating (which is apparently the con- 
dition in which our form is usually picked up), of mucous appearance, with a whitish 
fibrous centre composed of tubes coming from the ends of the animals ( = Ascidiozooids). 
It was found " 30 leagues from Terre-des-Etats," at the southern extremity of America. 
I notice that Michaelsen (Hamburger magalhaensische Sammelreise, " Tunicaten," 
1907, p. 40) has also suggested with a (?) that Lesson's Holozoa cylindrica is the same 
as " Julinia " (or Distaplia) ignota. 

Family Polyclinid^e. 

Polyclinum cornplanatum, Herdman (?). 

The species was described t from a specimen obtained at Port Jackson, Australia. 
The Scotia material was taken at Station 483, at the entrance to Saldanha Bay, on 
May 21, 1904, from a depth of 25 fathoms. It consists of four fragments, cut probably 
from the same colony, the largest of which measures about 6 cm. by 2. The colony 
was apparently flattened, and had much the same shape and colour as the Australian 
one. The Ascidiozooids also have the same type of structure. The post-abdomen is 
rather longer than in the Australian specimens, but that is a matter that varies with 
the reproductive condition. The specimens are, however, so fragmentary, and there 
is so little that is distinctive, that I cannot be certain as to the identity of the species ; 
but there is nothing in the microscopic details to negative the view that the Falkland 
Islands specimens belong to this Australian species. 

Amaroucium distomoides, Herdman (?). 

I refer one large colony and a few small fragments in the Scotia collection to this 
Australian species. J The original specimen came from Port Jackson; the Scotia 

* In the new edition of the " Tunicata" of Beonn's Tier-Reichs. 

t See Herdman, Descriptive Catalogue of the Tunicata of the Australian Museum, Sydney, N.S.W., 1899, p. 81. On 
the plate (Pel. I. figs. 9-12) it is referred to as "Polyclinum depressum." 
I See Herdman, ibid., p. 75. 



318 PROFESSOR W. A. HERDMAN ON THE 

material is from Port William, Station 349, Falkland Islands, February 6 to 8, 1903, 
6 fathoms. The large colony measures about 14 cm. by 3 cm., and is attached along 
the length of a Laminaria-like Alga. The test is dark greyish brown, and the small 
yellow Ascidiozooids show all over the surface as closely placed dots or streaks of a 
lighter colour. In further details this specimen agrees well with the description of the 
Australian one. One zooid was, however, noticed with an eight-lobed branchial aperture. 
The stomach has longitudinal folds. The stigmata are large. The dark colour of the 
test is due to dense crowding with small test-cells. The Scotia colony was evidently 
taken at the reproductive season, as it contains abundance of embryos in various stages 
of development up to the tailed larval stage ready to be set free. 

Amaroucium sp. (?). 

Some small colonies, a few millimetres to about 1 centimetre across, which were found 
attached to groups of Styela paessleri and other Ascidians from the Falkland Islands, 
belong to the genus Amaroucivm, but may be only young colonies of some larger form 
such as A. distomoides, or A. pallidulum obtained by the Challenger Expedition at 
Port William. 

It may be remarked in regard to the three last species of Compound Ascidians that 
they require re-examination in the living state. Many of the Compound Ascidians 
are scarcely determinable from the contracted and bleached specimens in preserved 
collections. It may well be that one or other of the above Polyclinids had in 
the living state a bright colour or some other characteristic appearance that is now 
wholly lost. 

THALIACEA. 

(MS. received March 13, 1912.) 

Family Salpid.33. 

The very large collections of Thaliacea, which were obtained at the South Orkneys 
and other Antarctic localities (some from under the ice), were found on examination 
to belong entirely to the genus Salpa and to represent two species only ; and in fact 
all the specimens, except a single one, are different conditions and sizes of the common 
and widely distributed species, Satya runcinata-fusiformis. 

Salpa runcinata-fusiformis, Chamisso-Cuvier. 

Station 432, surface, March 30, 1904 ; temp. 31 "8°. Nearly one hundred specimens, 
from 3 cm. to 6 cm. in length, all of the aggregated form, and many of the larger 
ones showing echinated ridges on the test. Most of them showed embryos projecting 
into the peri-branchial cavity, one in each. 

Station 427, from coarse tow-net, March 2G, 1904. About one hundred specimens, 
from 3 cm. to 5 cm. in length. In other respects they resemble those from the last 



TUNICATA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 319 

locality, except that the nuclear mass has more of a canary-yellow tint in these, and 
was a pinker colour in the others. 

Station 325, trap-hole, surface, June 4, 1903, Scotia Bay, South Orkneys. A 
dozen specimens, all rather large, up to 6 cm., and having some red pigment on 
the nucleus. 

Station 416, taken by trawl lowered to 2370 fathoms, but probably captured at 
surface, March 17, 1904, lat. 71° 22' S., long. 18° 15' W. About forty specimens, 
much smaller, up to 4 cm. in length at most, all with a pale yellow-coloured nucleus. 

Station 422, vertical net, surface to 800 fathoms, March 23, 1904 ; temp. 31*1° ; lat. 
68° 32' S., long. 12° 49' W. This jar contains a large matted mass of Salpae, Medusae, 
Macrurous Crustacea, and small fish, also a large species of Sagitta. It looks as if it 
had at one time become dry. The Salpae seem to be all of the aggregated form of this 
species, and are of medium size. 

Station 391, "Tunicate," water-bottle on sounding wire (depth of sounding, 2630 
fathoms), February 27, 1904. A single, very large specimen of the solitary form, 
fully 8 cm. in length, but rather damaged, with a chain measuring at least 4 cm. 

Surface otter trawl, February 24, 1904. One specimen, aggregated form. 

Station 325, "Doliolum," trap-hole, surface, June 1903, Scotia Bay, South Orkneys. 
About twenty specimens of the chain form. 

"Doliola?," while sounding, March 22, 1904. Three specimens. 

Station 325, surface, June 1903, temp. 28'9°, Scotia Bay, South Orkneys. About 
ten specimens. 

Station 325, surface, June 1903, temp. 29°, Scotia Bay, South Orkneys. Three 
specimens. 

Station 430, vertical net, March 28, 1904, temp. 31 *0° F. About fifty specimens, 
badly preserved, along with some Medusae. 

Station 432a, surface, March 30, 1904, temp. 31'8°. Haifa dozen specimens. 

Station 325, surface, June 1903, temp. 29°, Scotia Bay, South Orkneys. One 
damaged specimen. 

Station 409, "Tunicate," vertical net, fathoms 0-100, March 5, 1904, temp. 30°. 
One damaged specimen. 

Station 391, evening, while sounding, February 27, 1904. One damaged specimen. 

Salpa scutigera-confederata, Cuv.-Forsk. 

Station 535, on Gulf weed, surface, June 27, 1904, "Doliolum." One badly preserved 
specimen, which had probably been dried at one period, solitary form of the species. 



320 THE TUNICATA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 



ADDENDUM. 

As this paper was going to press I received from Dr Bruce a couple of tubes con- 
taining the following : — 

Polyzoa pictonis, subspecies patagonica, Michn. One poor specimen from shore, 
Port Stanley, Falklands, January 1903. Station 118. 

Lissamaroucium magnum, Sluiter. One colony about 3 "5 cm. in diameter, trawled 
from Station 346, 56 fathoms, December 1, 1903, Burdwood Bank. 

Amaroucium sp. (?). One colony about 3 cm. in diameter from same haul as the 
last species. Station 346. 



EXPLANATION OF THE PLATE. 



Fig. 1. Styela lactea, Herdman. Large barrel-shaped specimen. Nat. size. 

Fig. 2. Part of large colony of Holozoa cylindrica, Lesson, with three small specimens of Styela lactea 
attached. Nat. size. 

Fig. 3. Globular specimen of Styela lactea, showing the positions of the branchial and atrial apertures. 
Nat. size. 

Fig. 4. Anterior end of similar specimen, showing the branchial aperture. Nat. size. 

Fig. 5. One of the echinated spines of Styela lactea. Enlarged. 

Fig. 6. Small, globular and very spiny specimen of Styela lactea. Nat. size. 

Fig. 7. Alimentary canal of Styela lactea. Slightly enlarged. 

Fig. 8. Dorsal tubercle and tentacles of Styela lactea. x 40. 

Fig. 9. Dorsal tubercle of Paramolgula gregaria, Lesson. Enlarged. 

Fig. 10. Paramolgula horrida, Herdman, right side. Nat. size. 

Fig. 11. Part of branchial sac of P. horrida. x 40. 

Fig. 12. Diagrammatic plan of both sides of branchial sac of Styela paessleri, Michaelsen, supposed to 
be cut through the endostyle and spread out ; I. to IV., branchial folds. The number of bars on the folds and 
in the interspaces is shown. 

Fig. 1 3. Small part of branchial sac of Styela paessleri. x 40. 

Fig. 14. Dorsal tubercle of Styela paessleri. x 40. 

Fig. 15. Fungulus antarcticus, n.sp., from the left side. Nat. size. 

Fig. 16. Branchial aperture of the same. Enlarged. 

Fig. 17. Part of branchial sac of same species, from the inside. x 40. 

Fig. 18. Dissection of Fungulus antarcticus, to show alimentary canal and gonads. A little enlarged. 



i. Roy. Soc. Edin r 



Hekdman: "Scotia" Tunicata 



Vol. XLVIII. 




Ml'arlane & Erskino, Litli.. Edin. 



( 321 ) 



XVII. — Scottish National Antarctic Expedition : Observations on the Anatomy of 
the Weddell Seal (Leptonychotes Weddelli). By David Hepburn, M.D., CM., 
Professor of Anatomy, University College, Cardiff (University of Wales). Part III. 

(MS. received March 28, 1912. Bead June 3, 1912. Issued separately July 18, 1912.) 

The Respiratory System, and the Mechanism of Respiration. 

In the specimen under consideration there were fifteen pairs of ribs, of which nine 
pairs were vertebro-sternal. The costal cartilage associated with each of these was long 
and very flexible. The articulation of the first pair of costal cartilages with the sternum 
was effected by means of a short but strong band of fibrous tissue, which permitted 
considerable freedom of movement and did not form a junction of the more or less rigid 
character seen in man. 

The chondro-sternal joints of the second, third, and fourth costal arches were of the 
diarthrodial variety, each joint being divided into two separate cavities by an inter- 
articular ligament. The fifth, sixth, seventh, eighth, and ninth chondro-sternal joints 
presented diarthrodial joints without interarticular ligaments. 

The sternum was long, narrow, somewhat like a four-sided rod, and divided into 
segments (suggestive of vertebral centra) by amphiarthrodial joints which were 
placed opposite the chondro-sternal joints from the second to the ninth. A suprasternal 
tapering cartilage extended towards the head for a distance of two inches, while the 
ensiform cartilage extended backwards to a similar distance and ended in a broad semi- 
lunar expansion. 

The intercostal muscles were well developed, being thick and fleshy, presenting little 
or no, fibrous intersection. They were arranged so as to present an external and an 
internal muscle in each intercostal space, and the direction of their fibres was similar to 
that seen in man ; but the fibres of the external muscle were continued between the 
costal cartilages close up to the margin of the sternum without the intervention of an 
intercostal membrane. 

The triangularis sterni muscle arose from the deep surface of the sternum on its own 
side of the mesial plane. It consisted of a number of slips, which were wide enough to 
give the appearance of a complete sheet of muscle. These were attached to the sternum 
from the level of the third costal cartilage backwards to the level of the ninth. The 
fibres ran forward and outwards to be inserted into the deep surfaces of the costal 
cartilages from the second to the ninth inclusive, and into fibrous bands which passed 
from one cartilage to the other. The general line of insertion into the costal cartilages 
was near to the series of costo-chondral joints, each of which, except that of the first rib, 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 17). 49 



322 PROFESSOR DAVID HEPBURN ON 

formed a cliarthrodial joint. On the ninth costal cartilage this muscle interdigitated 
with the attachment of the diaphragm. This large, well-developed muscle was supplied 
by a series of twigs derived from the intercostal nerves in relation to which it was 
attached. 

The sterno-mastoid muscle extended from the anterior end of the sternum and from 
the side of its pointed suprasternal cartilage to the mastoid process. There being no 
clavicle, this muscle appeared narrow. 

The sterno-hyoid and sterno-thyroid muscles arose from the sternum under cover of 
the previous muscle. They formed a thin continuous sheet which probably included 
the omo-hyoid muscle along its lateral border in the vicinity of the hyoid bone. The 
entire sheet was innervated from the hypoglossal nerve, and the insertion of fibres into 
the thyroid cartilage and into the hyoid bone suggested the character of its constituent 
parts. A thin band of muscle fibres occupying their usual position formed the thyro- 
hyoid muscle. 

There were two well-defined scalene muscles, both of which were situated on the 
dorsal side of the subclavian vessels and cervical nerves, and may therefore be regarded 
as the representatives of the scalenus medius and scalenus posticus muscles. 

The musculus scalenus medius was inserted into the costal cartilage of the first rib 
close to the costo-chondral articulation. 

The musculus scalenus posticus was inserted into the lateral aspects of the third, 
fourth, fifth, and sixth costal cartilages close to the costo-chondral articulations. At 
each insertion the pointed attachment interdigitated with similar attachments of the 
musculus obliquus externus abdominis, whose digitations extended to the cartilage of 
the first rib. 

Regarding the skull and the cervical column as providing the more fixed or rigid 
attachment for the scalene and sterno-mastoid muscles, it is fairly evident that these 
muscles may act as elevators of the ribs and sternum by drawing them towards the head. 

The diaphragm was well defined in all its parts, but its dorso-lateral portions were 
very thin, and in the absence of a central tendon of the trefoil type it presented appear- 
ances deserving detailed description, more especially in regard to the important position 
occupied by this muscle in the mechanism of respiration. Its strongest part was the 
mesial or vertebro-sternal element, which presented two well-marked, pointed crura 
attached to the lumbar vertebras. From this origin the muscular fibres passed in a 
ventral direction on either side of the abdominal aorta until they reached the ventral 
aspect of this vessel, where to a small extent their fibres intermingled ; but for the most 
part the fibres of the right crus were on the ventral side of those of the left crus. 

This distinction between the fibres of the two crura was maintained as they continued 
towards the oesophagus, along the lateral aspects of which they passed, thereby forming 
the oesophageal opening, which was practically in the mesial plane. A short distance 
on the ventral side of the oesophageal opening the muscular fibres were inserted into a 
circular tendinous ring placed slightly to the right of the mesial plane, and through this 



THE ANATOMY OF THE WEDDELL SEAL. 323 

ring the inferior vena cava passed. From the ventral face of this tendinous ring strong 
muscular bands extended to the deep face of the broad ensiform cartilage. 

From each side of the dorsal segment of the fibrous ring surrounding the inferior 
vena cava there extended a narrow tendinous septum in the dorso-lateral direction. 
Neither of these septa reached the ribs, although the one on the right side was more 
strongly marked than that on the left side. Into the dorsal faces of these septa there 
were, inserted muscular fibres derived from the lateral aspects of their respective crura, 
as well as a small, feeble muscular slip from the ventral surface of the second last rib 
(the 14th) near its head. 

From the ventral aspects of the tendinous septa under consideration, a sheet of 
muscular fibres passed outwards to be attached by slips or digitations into the ribs close 
to their junction with cartilages from the 8th to the 1 3th inclusive. The cligitation be- 
longing to the 13th rib was attached just in front of the angle of the rib. This ventro- 
lateral part of the diaphragm was very thin. Areolar tissue occupied the intervals 
between digitations attached to the 13th and 14th ribs, and also between the 14th rib 
and the lateral margin of the crus. The association of the diaphragm with the 1 5th rib 
was so feeble as to be doubtful. Probably these weak places may be regarded as corre- 
sponding to arcuate ligaments, although in the human sense these structures were 
undefined. At any rate, these arched ligaments had no other representation, neither 
could the slight intermingling of crural fibres on the ventral aspect of the "aorta be 
regarded as other than a very feeble median arched ligament. Altogether the dorso- 
lateral development of the diaphragm was extremely feeble. Those muscular fibres 
attached to the 8th rib were in close contact with the mesial part of the diaphragm 
between the sternum and the fibrous ring enclosing the inferior vena cava. 

From what has been described it will be evident that there was no central tendon of 
the trefoil pattern, but in its place a vena caval ring, from which there extended two 
dorsal-lateral septa, of which the right was the stronger marked. The shortness of the 
left septum permitted a greater mingling of the fibres belonging to the left crus with 
those forming the rest of the left half of the diaphragm. 

The general arrangements within the chest cavity do not call for special discussion. 
The two pleural sacs and the mediastinal interval followed the customary disposition. 
It may, however, be noted that the mediastinal layers of pleural membrane were in close 
apposition from the second segment of the sternum backwards to the hinder end of the 
sternum, and that consequently the ventral or anterior section of the mediastinum was 
a mere chink, in its sternal relations. 

The lungs were extremely dark in colour, brown almost to black. They were quite 
soft, but yielded no feeling of crepitation on pressure, so that they suggested a complete 
absence of air. 

When they were removed each was placed in water, and they sank as if solid. A 
small portion cut from the lung also sank in water, so that this tissue had entirely lost 
its buoyancy. It is not quite easy to account for such a complete absence of air from 



324 PROFESSOR DAVID HEPBURN ON 

the substance of the lung. The animal is known to have lived for some days, because it 
was killed by poisoning with hydrocyanic acid after an attempt had been made to rear 
it by artificial feeding. The carcase was preserved by injecting an arsenical solution. 
Neither of these processes would account for the absence of air from the lung tissue. 
We must therefore assume that either the natural elasticity of the lung tissue has pro- 
duced the condition noted, so that the expiratory apparatus of the animal is able to pro- 
duce a practical deflation of the lungs, which is doubtful ; or that, partly owing to the 
length of time they have been preserved and partly owing to the preservative solutions, 
the air has practically all passed into solution and disappeared. 

A portion of the lung was prepared for microscopic examination, and, notwithstanding 
the number of years that have elapsed since the animal was embalmed, the different 
tissues were easily recognisable, but to staining agents such as hsematoxylin and eosin 
they reacted very slowly and not very satisfactorily. 

The hyaline cartilage of the bronchioles was cellular, and very similar to the cartilage 
in the ear of the mouse. 

The lobules of the lung were very clearly defined by interlobular tissue, which was 
continuous with the sub-pleural tissue, and throughout this tissue there was a well- 
marked amount of elastic fibres. 

All the air spaces were shrunken, i.e. collapsed, almost to the point of obliteration, 
but they were free from exudation. The capillary blood-vessels in the walls of the 
air spaces were crowded with blood corpuscles, which may have been the result of the 
preservative injection. 

There is some reason, therefore, for considering that the normal elasticity of the lung 
in this seal was much greater than that of man, and that, consequently, the air would 
be much more effectively expelled from the lungs of the seal during expiratory move- 
ments. 

Attention may be drawn to certain of the body muscles whose attachments and dis- 
position were such as to add to their expiratory value. The panniculus carnosus muscle 
was a thin sheet enveloping the trunk from the hinder end of the abdomen to the face, 
and on the face and head forming a cowl modified for facial or expression muscles in 
relation to the various apertures in that region. The fore limbs were in effect pushed 
through this axial sheet. The disposition of its fibres showed dorso-lateral and ventro- 
lateral directions, separated from each other by a lateral aponeurosis, and attached by 
aponeurotic fibres to the dorsal and ventral mesial lines such as may be seen in the 
porpoise, but less distinct. The direction of the muscle fibres in the dorso-lateral section 
was obliquely from before (cephalic) backwards (caudal), whereas in the ventro-lateral 
section their direction was obliquely from behind forwards. The general effect of the 
contraction of this sheet would be to expel the air from the very elastic and flexible 
thorax, as well as to compress the abdomen. 

The musculus obliquus abdominis externus showed no attachment to the ilium. 
By one end it was attached through digitations to the entire series of costal arches from 



THE ANATOMY OF THE WEDDELL SEAL. 325 

the first to the fifteenth. Its fibres were directed obliquely backwards towards the 
ventral mesial line, and, having given place to a thin aponeurotic sheet, many of these 
fibres interlaced with those from the opposite side to form the linea alba. 

The hinder part of the muscle, however, did not form any attachment to the ilium, 
but as a muscular arch, equivalent to the ligament of Poupart, they were attached to 
the ventral aspect of the body of the pubis. Near to the pubis a slit in the muscle 
sheet served the purpose of an external inguinal ring, in which the spermatic cord was 
situated. A muscle so attached could clearly act as a very powerful expiratory muscle 
provided the glottis were open. 

The musculus obliquus abdominis internus was attached dorsally to the lumbar 
aponeurosis and to the crest of the ilium, while ventrally it was inserted into the hinder 
borders of the last four ribs and also through its aponeurosis into the linea alba. The 
greater proportion of its aponeurotic fibres passed ventrally to the rectus abdominis 
muscle along with those of the external oblique, but a few of them blended feebly 
with the aponeurosis of the transversalis abdominis muscle. This muscle (transversalis) 
presented lumbar and iliac attachments as well as a series of digitations on the hinder 
seven or eight ribs. Its mesial attachment to the linea alba was by an aponeurotic 
sheet placed on the deep side of the rectus abdominis muscle. The hinder or inguinal 
margins of these two last muscles were very closely, almost inseparably, blended 
together, and both were much thinner than the external oblique muscle. 

The rectus abdominis muscle occupied an abdominal sheath whose composition has 
already been indicated. It was attached posteriorly to the body of- the pubis, and 
extended anteriorly to the first costal cartilage, to which, as well as to all the other 
sternal cartilages, it was attached by tendinous slips. Here again we can see that this 
muscle, acting from a rigid attachment to the pubis, may act as a powerful expiratory 
muscle in association with an open glottis. 

The lungs, beyond what has already been said, do not call for detailed description. 
Each presented a great oblique fissure, and thereby an apical and a basal lobe. In 
addition the right lung possessed a transverse fissure, and therefore a middle or ventral 
lobe. Furthermore, the right lung had an azygos lobe on its mediastinal aspect in 
relation to the margin between diaphragm and pericardium. 

On several occasions I have had the opportunity of making a detailed examination 
of the respiratory mechanism of mammals whose habitat is either partly or entirely 
marine, and on each occasion I have been impressed by the remarkable flexibility of 
their thoracic wall, with the associated peculiarities in the attachments of certain of the 
muscles. Attention has already been drawn to some of these peculiarities in the 
descriptions given above, and it is almost impossible to avoid the conclusion that 
respiration necessitates a more flexible chest-wall in the case of mammals surrounded 
by water than in those which are surrounded by air, apart from the fact that the 
normal attitude of the latter may be horizontal, as in the case of quadrupeds, or vertical, 
i.e. erect, as in the case of man. 



326 PROFESSOR DAVID HEPBURN ON 

On the other hand, this flexibility of the chest becomes not only a drawback, but 
may be an actual source of danger when the marine mammal comes on shore either by 
intention or as the result of accident. Thus it is well known that a cetacean dies when 
it runs aground, not necessarily by starvation, but because it is suffocated, since the 
flexibility of its chest-wall renders respiratory movements impossible under the superin- 
cumbent weight of its body. 

The leader of the Scottish National Antarctic Expedition made very careful observa- 
tions on the attitude of seals when they left the water and resorted to the ice, and he 
noted that they do not assume positions which would hamper their chest movements. 
Thus they recline on the side when asleep so as to leave the movements of one side of 
the chest unimpeded, while at other times their common attitude is to lie prone with 
the chest raised off the ground by the short fore limbs. Considerations of this kind lead 
to the conclusion that respiration, but more especially the act of inspiration, can be 
seriously impeded or even rendered impossible by the weight of the animal's own body. 
That free inspiratory movement of the chest-wall in man may be hampered by the 
weight of his body may be readily observed in the case of an operatic singer who, by 
the exigencies of his performance, is called upon to sing in the supine position ; for 
although in this position he can fill his lungs sufficiently for ordinary respiration, 
yet he cannot inspire deep enough for effective vocalisation. Clearly, therefore, the 
respiratory mechanism is affected by the attitude of the individual as well as by the 
surrounding medium, air or water, in which the animal performs the necessary 
respiratory movements. 

There can be no doubt that, whatever the natural attitude of the mammal may be, 
or whatever its habitat, the ordinary movements of inspiration and expiration are carried 
out with the minimum expenditure of effort consistent with the amount of air required 
for each respiratory act. On the other hand, special circumstances may call for 
additional or extraordinary efforts both as regards inspiration and expiration. The 
discussion of respiratory movements is usually left, and by many observers considered 
properly left, to the physiologist ; but as these movements are entirely dependent 
upon a definite mechanism in which the muscular arrangements play an important 
part, they cannot fairly be excluded from the province of the anatomist, and it is 
from the standpoint of structure that I propose to offer some observations which 
seem warranted by the conditions I have seen in the seal under consideration, as well 
as in the porpoise. 

It may be well in the first instance to deal with the lungs themselves ; and, as the 
condition in which I found them has already been stated, it will only be necessary to 
add that, except for the presence of the azygos lobe on the right side, they corresponded 
with the human lungs so far as the number and arrangement of fissures and lobes was 
concerned. There is no reason to suppose that during the act of inspiration they would 
inflate in a manner different from the lungs of man. Now, among the many interesting, 
elaborate, and ingenious attempts to explain the respiratory act, none is more suggestive 



THE ANATOMY OF THE WEDDELL SEAL. 327 

than that of Keith,* who approaches the discussion of the subject more as an anatomist 
than as a physiologist. 

It is not my intention to follow Professor Keith in detail and offer a criticism of the 
conclusions arrived at by him. At the same time, some of his statements appear to me 
to overlook certain of the anatomical facts. Perhaps the most important fundamental 
statement made by Keith is in reference to the lungs when he says " the upper lobe is 
normally expanded by one mechanism, the lower by another," and as a consequence he 
insists that " the great fissure, which divides the upper from the lower lobe, is functional 
in its significance." Supposing this view to be correct, it would follow that since there 
is a third lobe in the right lung of man and a fourth or azygos lobe in the right lung of 
a quadruped, with the fissures required for their delimitation, the mechanism for expand- 
ing the right lung must differ from that required for the left lung. Further, as regards 
the apical or upper lobe, Keith maintains that because of the impressions of certain ribs 
upon the lateral and anterior aspects of the upper lobe, but not upon " the dorsal surface 
of the upper lobe," there is " a constant relationship between ribs and spaces " for that 
part of the lobe which presents impressions, but a " downward and upward " movement 
of the dorsal unmarked part, in which it follow^ the movement of the lower lobe, because 
" the lower lobe and the dorsal part of the upper lobe are chiefly expanded by a dia- 
phragmatic mechanism." The argument for a functional significance for the great oblique 
fissure seems to me unnecessary if the substance of the apical lobe is to expand in two 
different ways simultaneously, for, at least as far as the dorsal part is concerned, the 
presence of the fissure does not seem to confer any advantage. 

I am not disposed to maintain that the fissures of the lungs have no significance, 
although to my mind it is rather structural than functional. Even " the obliteration 
of the pleural cavity by adhesions has so little apparent effect on the respiratory 
movements that their presence cannot be detected during life," any more than the 
obliteration of the lobulated character of the kidneys interferes with their functions. 
After all, the outstanding requirement is that the lungs shall expand to the capacity 
corresponding to the immediate muscular effort that is being performed, and naturally, 
therefore, the capacity undergoes constant variation. With this end in view, I cannot 
but think it is best to consider the muscular mechanism of inspiration as a whole, and 
the muscular mechanism of expiration as a ivhole, since it is their co-ordinated and not 
their individual action that we depend upon. Probably, in quiet ordinary breathing, 
no animal, any more than the human individual, employs the full scope of its inspiratory 
mechanism, and hence in man it has become customary to employ such terms as 
"thoracic" and "abdominal" to indicate the character of the inspiratory effort which 
is most noticeable in the female and in the male respectively. At the same time, there 
is no record of this distinction in the inspiratory act among the sexes of the lower 
animals, nor between the human sexes during infancy and early adolescence. It 

* "The Mechanii-.m of Kespiration in Man," by Arthur Keith, pp. 182-207, in Further Advances in Physiology, 
edited by Leonard Hill, published by Edward Arnold, London, 1909. 



328 PROFESSOR DAVID HEPBURN ON 

appears to me, therefore, that the double mechanism which Keith supports for the 
expansion of the apical and basal lobes of the lung is more apparent than real, and that 
the so-called types of breathing in man are rather the result of the erect attitude 
whereby, from the natural configuration and diameters of the diaphragmatic section of 
the thoracic cavity in the two sexes, it is with less muscular effort that the male expands 
the lower part of his chest and the female expands the upper part of her chest in order 
that each may obtain the amount of air necessary for ordinary quiet breathing. When 
additional efforts call for more air, or when, as in the supine position, the easy movement 
of the chest is impeded, then there is an immediate departure from the characteristic 
method ; but I do not think that we must postulate a double inspiratory mechanism. 

The key to the whole mechanism of inspiration is undoubtedly the part performed by 
the contraction of the diaphragm. We cannot, therefore, overestimate the importance 
of its attachments and structure ; nor must we forget that, like any other muscle, its 
action is the result of the contraction of its fibres, whereby its attached ends are brought 
more or less near to each other. The most favourable method of examining the diaphragm 
is to consider the adult structure from the point of view of its development. 

The first part of the diaphragm to appear in the embryo, and the part which may 
be considered the most powerful in the adult, is its mesial or vertebro-sternal portion, 
whose vertebral ends or crura arise from the lumbar vertebrae and constitute its axial 
or fixed end. These muscular fibres having adapted themselves to the positions of the 
abdominal aorta and the oesophagus, by a certain amount of intermingling of the fibres 
from opposite sides of the mesial plane, become inserted into the central tendon, whose 
shape varies from the trefoil tendon of man to the vena caval ring with its lateral septa 
as seen in the Weddell seal. From the ventral aspect of these tendinous structures a 
second set of muscular fibres extends to the deep surface of the lower or hinder end of 
the sternum. The mesial part of the diaphragm is therefore in reality a digastric 
muscle pursuing an arched course from the vertebral column to the sternum. The 
arched character of its course is more pronounced in its dorsal segment, while in its 
ventral or sternal segment the arched character is lost, being replaced by a straight or 
flat course. This change in the curve of the two segments is due partly to the disposi- 
tion of the abdominal viscera and partly because the vertebral attachments are some 
distance farther tailwards than a point which would correspond with the hinder end of 
the sternum. 

By its contraction two results may follow : — (1) the lower (hinder) end of the sternum 
is either drawn closer to the vertebral column or else prevented from being projected 
ventrally (forwards) by other influences; (2) the arched dorsal segment between the 
vertebral column and the central tendon becomes more or less flattened, and in conse- 
quence the adjacent abdominal viscera are pushed towards the ventral abdominal wall. 
At the same time its restraining or bracing action upon the hinder end of the sternum 
becomes correlated to the restraining action of the first costal arches upon the manubrium 
sterni, whereby the manubrium sterni is maintained at a relatively fixed distance from 



THE ANATOMY OF THE WEDDELL SEAL. 329 

the vertebral column. By reason of the somewhat rigid character of the first costal 
cartilages in man, as well as from the fact that they are frequently encased in an ossified 
shell, the restraining nature of the connection between the first pair of ribs and the 
sternum is very well marked ; but even in the Weddell seal, where a short and powerful 
fibrous ligament takes the place of the first costal cartilage, the manubrium is very 
firmly retained in its relation to the backbone. 

The next part to be added to the diaphragm developmentally constitutes its ventro- 
lateral segments. In the adult these are composed of muscular fibres which arise from 
the ventral and lateral aspects of the central tendon. From this position they extend 
in a fan-shaped manner to be inserted into the deep surfaces of the costal arches by 
digitations which correspond very closely in number with those ribs that do not reach 
the sternum directly through their costal cartilages — that is to say, the false or vertebro- 
abdominal series of ribs. 

This thin sheet of muscle becomes more and more arched as its slips sink lower on 
the series of ribs. When therefore it contracts, each digitation will either draw its own 
particular rib nearer to the central tendon or else maintain the ventral end of its rib 
at a more or less definite distance from the central tendon. 

In this way the ventral ends of the false ribs are provided with temporary or inter- 
mittent fixed points, fixation by the contraction of the diaphragm being substituted 
for fixation by the sternum, as is the case with vertebro-sternal ribs. In fact, the series 
of ribs could with effect be classified as vertebro-sternal and vertebro-diaphragmatic. 

The flattening of the arched surfaces of the diaphragm must increase the available 
thoracic space, but under ordinary conditions the addition so provided cannot of itself 
be very great, and only becomes important as the central feature of a larger movement. 

Developmentally, the last part to be added to the diaphragm is also its weakest 
part both in man and in the Weddell seal. This is the dorso-lateral segment, which 
consists of muscular fibres forming a delicate sheet extending between the dorso-lateral 
aspects of the central tendon and the ligamenta arcuata externa and interna, and 
through these with the vertebral column on the one hand and the last rib on the other. 
The arched course of these fibres in man must enable them to aid the flattening of the 
dorsal parts of the diaphragm and thereby again assist in pushing the abdominal con- 
tents in a ventral direction, but in the seal they are so feebly developed that the effect 
of their contraction must be practically negligible. I do not doubt that contraction of 
the diaphragm may produce some depression of the central tendon, more especially at 
its dorsal side, but I doubt whether the depression of the central tendon can take place 
on its sternal side or be so pronounced as a whole as to give the " piston action " 
described by some observers. My reasons for holding this view may be shortly 
summarised. The pericardial bag rests by its base upon the diaphragm, and, when a 
central tendon of the trefoil pattern is present, the fibrous bag and the central tendon 
are intimately united to each other, but the ventral surface of the pericardium is 
attached to the manubrium sterni by a sterno-pericardial ligament which is described 

TRANS. ROY. SOC. EDIN., VOL. XLVIIL, PART II. (NO. 17). 50 



330 PROFESSOR DAVID HEPBURN ON 

by Macalister * as strong and rounded. A similar ligament of weaker character 
attaches the pericardium to the ensiform cartilage. These ligaments, especially the 
former, would resist the traction of the pericardium in the abdominal direction, and so 
resist the depression of the central tendon. Again, the normal liver presents indented 
grooves corresponding to, and resulting from apposition with, the ribs which cover it, 
and these grooves indicate a fairly constant relation between the liver and the ribs, 
since they could not be formed by any plunging or piston action communicated to the 
liver from the diaphragm. If, on the other hand, the flattening of the dorsal portion of 
the diaphragm pushed the liver ventrally towards the ribs, and at the same time the 
contraction of the ventro-lateral portions of the diaphragm drew the lower ribs towards 
the liver or even maintained them in a position to resist the liver, then the liver mark- 
ings would be at once accounted for. Further, Starling t states that during ordinary 
respiration the central tendon of the diaphragm is practically motionless, but that as 
soon as respiration becomes laboured there is an actual downward movement of the 
diaphragm, and that during laboured inspiration the breathing is mainly thoracic in 
both sexes, "and the abdomen recedes with each inspiration." Apparently, therefore, 
according to this observer, it is fair to conclude that a laboured inspiration, by calling 
for more powerful action of the diaphragm, results not only in greater flattening and 
depression of the dorsal segment of the arched diaphragm, but also in the lower ribs 
being drawn closer to its central tendon than during ordinary breathing, which is just 
what an examination of the muscular attachments would lead one to expect. I there- 
fore arrive at the general conclusion that the diaphragm is the keystone in the inspira- 
tory mechanism, and that its chief action consists in resisting the ventral (forward) 
movement of the hinder (lower) end of the sternum and of those ribs which are not 
directly articulated to the sternum. As a result of this controlling action the whole 
series of ribs may participate uniformly in a general lifting or elevating movement 
which characterises their position at the end of inspiration as contrasted with then- 
sunken or depressed position at the end of expiration. Such an elevation of the ribs 
does not call for any rotation of their shafts, and indeed, from the nature of their 
capitular and tubercular articulations with the vertebral column, rotation of the shaft 
would be impossible. But the chondro-sternal as well as the articulations just 
mentioned are from the nature of their ligaments adapted to the movements of eleva- 
tion and depression of the rib as a whole, and even a small amount of such movement 
at the vertebral end of a rib would tend to be magnified by the length and obliquity 
of its shaft. In fact, the capsules of the costo-transverse articulations are sufficiently 
long to permit of the gliding action necessarily associated with such elevation and 
depression. 

In the seal there are two muscles whose attachments are only readily comprehen- 
sible when considered as part of the mechanism for depressing the ribs after they have 

* Macalister, Text-book of Human Anatomy, 1889. 

t E. H. Starling in Schiifer's Text-book of Physiology, vol. ii. pp. 276 and 280. 



THE ANATOMY OF THE WEDDELL SEAL. 33] 

been elevated. These are the musculus obliquus externus abdominis and the musculus 
rectus abdominis. Both of these muscles are attached to the pubis, i.e. to the un- 
yielding or rigid pelvis, and between them they provide slips or digitations of insertion 
into nearly all of the ribs, even extending to the first. Further, the musculus obliquus 
abdominis externus has no attachment to the ilium — in other words, both of these 
powerful muscles were pubo-costal in their attachments. With a distended chest and 
a glottis firmly closed so as to render the chest wall fairly rigid, these muscles by their 
contraction could clearly compress the abdominal contents ; but in association with an 
open glottis, of necessity they must pull their rib attachments towards the pubis — in 
other words, they must act as depressors of the ribs and thus as powerful expiratory 
muscles. Such a depressor action compels one to presume and to accept an elevated 
position of the ribs during inspiration. 

There is nothing in the mechanism which seems to require one mode of action for 
inspiration by the vertebro-sternal ribs and another mode of action by the vertebro- 
abdominal ribs. Of course, the first costal arch, from the nature of its sternal articula- 
tion, is, even in the seal, capable of less elevation than those ribs whose shafts and 
costal cartilages are longer and whose sternal articulations permit greater freedom of 
movement ; but, in order to secure a uniform method of elevation throughout the series 
of ribs, it is necessary to provide some more rigid line or point d'appui for the start of 
the movement, and the combination of the first costal arch with the manubrium sterni, 
together with their powerful scalene and sterno-mastoid muscles, provides such a line. 
Moreover, in man the clavicle is added to this line by the sterno-mastoid and trapezius 
muscles as well as by such ligaments as the costo- clavicular and the sterno-clavicular. 
Again, it will be found that those ribs which elevate most readily are just those whose 
heads are provided with an interarticular ligament passing between the head and the 
intervertebral disc. Such a powerful structure would be unnecessary unless there 
were a tendency for the head of the rib to be drawn away from the vertebral column as 
the dorso- ventral and transverse thoracic diameters increased owing to the elevation of 
the ribs in full inspiration. There is no ordinary form of inspiration which requires a 
larger amount of air under regulated control than the inspiratory movement performed 
by the trained singer, and one of the approved methods of obtaining this result aims at 
the expansion of the large dorsal surface of the lungs by cultivating the upward move- 
ment of the ribs in relation to the dorsal surface of the chest, as being not only easier 
than, but preferable to, a forced action of the diaphragm. An examination of the 
mechanism of respiration leads me to the conclusion that, in all forms of respiration, 
this mechanism acts in the same way, but not, in all its parts, to the same extent, for 
any particular respiratory act ; that ordinary and laboured respiration differ in degree 
rather than in kind ; that the degree of respiration depends upon the amount of air 
required for any particular form of exertion ; that the difference between the respiration 
of the quadruped and the respiration of man results from differences in their attitude 
(horizontal and erect), whereby each cultivates that form of chest movement which 



332 THE ANATOMY OF THE WEDDELL SEAL. 

requires the minimum of muscular effort. The differences between the adult human 
male and female types of breathing are adaptations due to the avoidance of severe 
muscular effort so long as a smaller effort will serve the purpose, and in my opinion 
they result from the normal differences in the lower or diaphragmatic diameters of the 
thoracic cavity in the two sexes. To a large extent these differences may be accounted 
for by the width of the female false pelvis as compared with that of the male. Since 
the abdomen proper (i.e. excluding the true pelvis) contains no organs which are not 
common to both sexes, it follows that a wide false pelvis, by providing increased 
accommodation in the lower abdominal regions, is naturally associated with a reduction 
in the dimensions of the upper or diaphragmatic end of the abdomen. These conditions 
make elevation of the sternal ribs more necessary in the female than in the male, whose 
larger diaphragmatic thorax permits of ordinary breathing without the pronounced or 
visible elevation of his sternal ribs, although their movement may become visible when- 
ever the supply of air required calls for an extension of the elevating movement. 

In either sex, a change from the erect attitude to the horizontal (e.g. to the supine) 
position is usually followed by the introduction of more or less of the respiratory 
features of the opposite sex, owing to the temporary interference with the amount of 
rib movement in common use when the ribs are unobstructed. 



( 333 ) 



XVIII. —The Marine Mollusca of the Scottish National Antarctic Expedition. By 
James Cosmo Melvill, M.A., D.Sc, F.L.S., and Robert Standen, Assistant 
Keeper, Manchester Museum. Communicated by Dr W. S. Bruce. (With 
One Plate.) 

(MS. received April 24, 1912. Read June 3, 1912. Issued separately August 26, 1912.) 

PART II. 
Being a Supplementary Catalogue. 

Since we had the pleasure of working out the Mollusca obtained by the Scottish 
National Antarctic Expedition, Dr W. S. Bruce has kindly transmitted to our care 
some additional material, overlooked in the first instance, and taken (a) from deposits 
from jars in which Sponges were placed ; (b) from Algse and other growths, principally 
coming from Scotia Bay ; and (c) from a new species of Cephalodiscus. 

Of these the last, when macerated out and closely examined, produced the most 
prolific and interesting results ; but, notwithstanding this fact, the condition of many 
of the specimens extracted leaves much to be desired, so fragmentary and useless for 
scientific purposes was a very large proportion found to be. A certain few, however, 
are happily in better condition and recognisable, and, of these, we find several to have 
been described by Dr Hermann Strebel of Hamburg in 1908, the year subsequent 
to our first paper upon the subject being published. 

Others remain, of which over twenty do not appear to be represented in the collec- 
tions to which we could obtain access, nor mentioned in any of the treatises yet published 
on the Antarctic fauna. We are therefore emboldened to consider them new to science 
in the accompanying supplementary catalogue. 

We include afresh in the list of species obtained by this expedition those already 
catalogued in our first paper, thus rendering it as complete as possible, and signalise 
with an asterisk (*) those which are amongst the addenda now chronicled. 

We would thank Mr Edgar Smith, I.S.O.. for having examined some of the material, 
and likewise would express our indebtedness to the Rev. Lewis J. Shackleford, Messrs 
B. R. Lucas and J. Wilfrid Jackson, F.G.S., for having aided us in the difficult task 
of extracting such small and fragile objects from the mass in which they were too often 
almost hopelessly embedded. Mr T. Iredale has also kindly drawn up the description 
of a new species of Chsetopleura for this paper. 

We would only add that we have extended the Bibliographical Catalogue of the 
Antarctic Molluscan Fauna from 1907 to 1912 at the end of this enumeration. 

TRANS. ROY. SOC. EDIN., VOL. XLVIIL, PART II. (NO. 18). 51 



334 DR JAMKS COSMO MELVILL AND MR ROBERT STANDEN ON THE 

REVISED LIST OF SPECIES CONTAINED IN THE 
" SCOTIA " COLLECTIONS. 

A. Regio Antarctica — including Gough Island. 

Class GASTEROPODA. 

Order Amphineura. 

Sub-order polyplacophora. 

Callochiton illuminatus (Reeve). 
Tonicia atrata (Sowb.). 

# Plaxiphora setigera (King). 

# Chsetopleura brucei, Iredale, sp. n. 

# Hemiarthrum setulosum, Carpenter. 

# Lepidopleurus pagenstecheri, Pfeffer. 

Order Prosobranchiata. 

Sub-order diotocardia. 

(a) Docoglossa. 

Family Acmseidse. 
Acmsea ceciliana, Orbigny. 

Family Patellidse. 
Patella senea, Martyn, var. deaurata, Gmelin. 
,, fuegiensis, Reeve. 
,, polaris, Hombron and Jacquinot. 

(b) Rhipidoglossa. 
Section Zygobranchiata. 

Family Fissurellidse. 
Fissurella oriens, Sowb. 
„ picta, Gmel. 
Tugalia antarctica, M. and St. 

# Puncturella noachina (L.) 

Family Pleurotomariidw. 

# Scissurella eucharista, sp. n. 

„ euglypta, Pelseneer. 

,, mpraplicata, Smith 

,, tiraora, sp. n. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 335 

Section Azygobranchiata. 

Family Cyclostrematidse. 

* Cyclostrema calypso, sp. n. 

* ,, coatsianum, sp. n. 

* ,, gaudens, sp. n. 

* ,, meridionale, sp. n. 

Family Trochidse. 

* Calliostoma modestulum, Strebel. 
Photinula expansa (Sowb.). 

„ tseniata, Wood. 
,, violacea, King. 
Valvatella antarctica (E. Lamy). 

Sub-order monotocardia. 
Section (a) Ptenoglossa. 

Family Ianthinidae. 
Ianthina exigua, Lamarck. 

Family Scalidas. . 

* Scala magellanica, Phil. 

Section (b) Tsenioglossa. 

Family Naticidte. 

* Natica impervia, Phil. 

,, (Lunatia), sp. juv. 

Family Trichotropidas. 

* Trichotropis antarctica, sp. n. 

Family Capulidse. 

* Calyptrsea chinensis, L. 

,, costellata, Phil. 

„ dilatata, Lamk. 

Family Littorinidse. 
Littorina (Lxvilitorina) caliginosa (Gould). 
,, ,, coriacea, M. and St. 

,, (Pellilitorina) pellita, v. Marts. 
,, „ setosa, Smith. 

* Lacuna abyssicola, sp. n. 



336 



DR JAMES COSMO MKLVIIX AND MR ROBERT STANDEN ON THE 



Lacuna divaricata, Fabricius. 
,, notorcadensis, M. and St. 
,, ivandelensis, E. Lamy. 

Family Rissoidse. 
Rissoa adarensis, Smith. 

(Cingula) cingillus (Mont.). 
deserta, Smith. 
edgariana, M. and St. 
(0?ioba) jilostria, sp. n. 
fraudulenta, Smith. 
(Onoba) fuegoensis, Strebel. 
parva (Da Costa). 
( Onoba) paucilirata, sp. n. 

„ scotiana, M. and St. 

,, sulcata, Strebel. 
(Ceratia) turqueti, E. Lamy. 
(Manzonia) zetlandica (Mont.). 
Eatoniella kerguelenensis, Smith. 

Family Litiop>idse. 
Litiopa melanostoma, Rang. 

Family Cerithiidee. 
Cerithium georgianum, Pfeffer. 

,, pulhtm, Phil. 
Bittium brucei, sp. n. 

,, burdwoodianum, sp. n. 

Cerithiopsis macroura, sp. n. 

malvinarum (Strebel, MS.), M. and St. 

Family Turritellidse. 

* Turritella algida, sp. n. 

* Mathilda rhigomaches, sp. n. 

Family Tritoniidse. 
Gyrineum vexillum (Sowb.). 

Section (c) G-ymnoglossa. 

Family Eulimidse. 

* Eulima antarctica, Strebel. 

Family Pyramidellidse. 

* Turbonilla smithii, Pfeffer. 

„ xenophyes, sp. n. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 337 

Section (d) Rachiglossa. 

Family Muricidm. 

Trophon brucei, Strebel. 

,, cinguliferus, Pfeffer. 
,, crispus (Couthouy). 

* ,, falklandicus, Strebel. 
,, geversianus (Pallas). 
„ hoylei, Strebel. 

,, liratus (Couthouy). 

,, minutus (Strebel, MS.), M. and St. 

,, philippianus, Dunker. 

* Antistreptus magellanicus, Dall. 

Family Nassidse. 
Nassa (llyanassa) Vallentini, M. and St. 

Family Buccinidee. 
Chrysodomus (Sipho) archibenthalis, M. and St. 
,, ,, crassicostatus, M. and St. 

Neobuccinum eatoni, Smith. 
Euthria fuscata (Brug.). 
,, magellanica (Phil.). 
,, michaelseni, Strebel. 

* ,, rosea, Hombron and Jacquinot. 

Family Volutidw. 
Voluta (Cymbiola) ancilla, Solander. 
Guivillea alabastrina, Watson. 

* Mitra ( Volutomitra) porcellana, sp. n. 

Section (e) Toxoglossa. 

Family Conidx. 

Columbarium benthocallis, M. and St. 
Mangilia costata (Donovan). 

* Bela anderssoni, Strebel. 

* „ fulvicans, Strebel. 

* ? Thesbia, sp. 

* Savatieria concinna, sp. n. 



338 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Family Cancellariidw. 
Admete magellanica, Strebel. 

* „ limneseformis, Smith. 

* Paradmete typica, Strebel. 

Order Opisthobranchiata. 

Sub-order tectibranchiata. 

Section Bulloidea. 

Family Tornatinidse. 

* Retusa antarctica, sp. n. 

,, truncatula (Brug.). 

Section Siphonarioidea. 
Family SiphonariidsB. 
Siphonaria redimiculum, Reeve. 

Order Pulmonata. 
Sub-order basommatophora. 
Family Auriculidse. 
Marinula nigra, Phil. 

Class SCAPHOPODA. 
Dentalium eupatrides, M. and St. 
,, megathyris, Jousseaume. 

Class PELECYPODA. 
Order Protobranchiata. 
Family Nuadidse. 
Nucula minuscula, Pfeffer. 

* „ pisum, Sowb. 
Yoldia eightsii (Couth. ). 

* ,, profundorum, sp. n. 

Order Filibuanchiata. 
Sub- order anomiacea. 
Family Anomiidse. 
Anomia ephippium, Linn. 

Sub-order arcacea. 
Family Arcidee. 
Area (Bathyarca) strebeli, M. and St. 
Lissarca notorcadensis, M. and St. 

,, rubrofusca, Smith. 
Limopsis longipilosa, Pelseneer. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 339 

Sub-order mytilacea. 
Family Mytilidse. 
Mytilus edulis, Linn. 

,, magellanicus, Chemnitz. 
,, ovalis, Lamarck. 
Philobrya meridionalis (Smith). 
,, quadrata (Pfeffer). 

* „ sublwvis, Pelseneer. 

,, wandelensis, E. Lamy. 

* Crenella decussata (Mont.). 
Modiolarca mesembrina, M. and St. 

Order Pseudolamellibranchiata. 

Family Pectenidse. 
Pecten colbecki, Smith. 

,, multicolor, M. and St. 

,, ? patagonicus, King. 

,, pteriola, M. and St. 
Amussium 18-liratum, M. and St. 

Family Limidse. 
Lima goughensis, M. and St. 
„ (Limatula) pygmma, Philippi. 

Order Eulamellibranchiata. 
Sub-order submytilacea. 

Family Cavditidse. 

* Cardita congelascens, sp. n. 

* „ pallida, Smith, var. 1%-costata nov. 

Family Astartidse. 

* Astarte magellauica, Smith. 

Family Lucinidw. 

* Diplodonta lamellata, Smith. 
Cryptodon falklandicus, Smith. 
Cyamium antarcticum, Philippi. 

* ,, denticulatum, Smith. 

,, falklandicum, M. and St. 

Family Erycinidse. 
Laseea consanguinea (Smith). 
Kellyia cycladiformis, Desh. 



340 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Kellyia lamyi, M. and St. = australis, Lamy non Desh. 

* ,, magellanicus, Smith. 

* Davisia cobbi, Cooper and Preston. 
Scacchia plenilunium, M. and St. 

Sub-order tellinacea. 
Family Tellinidse. 
Tel Una (Msera) pusilla (Philippi). 

Sub-order veneracea. 
Family Veneridse. 
Chione philowela (Smith). 
Tapes {Amygdala) exalbida (Chem.). 

Sub-order myacea. 

Family GlycimeridsB. 

Saxicava arctica (L.), var. antarctica. Philippi. 

Sub-order anatinacea. 
Family Lyonsiidse. 
Lyonsia cuneata (Gray). 

Family Anatinidse. 
Anatina elliptica, King and Broderip. 

Order Septibranchiata. 
Family Cuspidariidse,. 
Cuspidaria brucei, M. and St. 

At Dr Bruce's request, we also include in the list of Mollusca obtained by the 
expedition certain species from St Vincent, Cape de Verde Islands, Pyramid Point, 
Ascension Island, and Funchal, Madeira. None of these call for special remark, 
beyond the fact that several, e.g. Area bouvieri, are endemic species, and that, so 
far as we can ascertain, Calliostoma montagui and Pisania maculosa have not been 
hitherto recorded from Cape de Verde. 

A. From St Vincent, Cape be Verde Islands. 

Chaetopleura fulva (Wood). 
Patella plumbea, Lamarck. 
Fissurella grseca (Linne). 
Haliotis lamellosa, Lamarck. 
Monodonta articulata, Lamarck. 
,, punctulata, Lamarck. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 341 

Monodonta turbinata (Born.). 

,, tamsi (Dunker). 

Calliostoma montagui (W. Wood). 

,, zizyphinus (Linne). 

Phasianella pulla, Linne, var. 
Pachypoma (Bolma) rugosa (Linne). 
Natica intricata, Donovan. 
Calyptrxa sinensis, Linne. 

,, (Infundibulum) radians, Lamarck. 

Littorina punctata, Gmelin. 

,, striata, King. 
Cerithium musicum, Sowerby. 

,, vulgatum, Brug. 
Planaxis lineatus, Cost. 
Cypraea spurca, Linne. 

Trivia arctica, Solander, var. europsea, Mont. 
Cassis testiculus, Linne. 
Obeliscus terebellum, Mull. 
Murex rosarium, Chem. 
Ocinebra corallina, Scacchi. 
Purpura hsemastoma, Linne. 

,, neritoidea, Linne. 
Collumbella, rustica, Linne. 

,, ,, var. azorica, Drouet. 

Nassa cornicula, Olivier. 
,, reticulata, Linne. 
,, cuvieri, Payr. 
Pisania maculosa, Lamarck. 
Leucozonia triserialis, Lamarck. 
Conus genuanus, Linne\ 
,, guinaicus, Brug. 
,, mediterraneus, Brug. 
Tethys punctata, Cuvier. 
Hammea navicula, Da Costa. 
Siphonaria venosa, Reeve. 
Area bouvieri, Fischer. 
Barbatia afra, Gmelin. 
Pectunculus formosus, Reeve. 

,, concentricus, Dunker. 

Mytilus puniceus, Lamarck. 
Pinna rudis, Linne\ 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 18). 52 



342 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Lithodomus aristatus, Dillwyn. 
Spondylus gadder opus, Lirm6. 
Cardita senegalensis, Reeve. 

ajar, Brug. 
Lucina pecten, Lamarck. 
Chama senegalensis, Reeve. 
Venus gallina, Linne. 
Chione nodosa, Dunker. 
,, verrucosa, Linne. 
Cardium edule, Linne. 

B. From Ascension Island. Pyramid Point, 40 fathoms. 

Nerita ascensionis, Gmelin. 
Pecten miniaceus, Reeve. 
CJiama, sp. 

C. From Funchal, Madeira. Shore. 
Patella cserulea, Linne. 



Class GASTEROPODA. 

Order Amphineura. 
Sub-order polyplacophora. 

Chiton (Plaxiphora) setiger, King 

Chiton setiger, King, Zool. Journ., v. p. 338 (1831). 
,, „ Sowerby, Conch. Ittustr,, p. 17. 

,, „ Zool. Beechey's Voyage, pi. xl. fig. 7. 

,, „ Reeve, Conch. Icon., t. ix. fig. 48a; t. xiv. fig. 48c. 

„ Gould, U.S. Explor. Exped. : Moll, p. 330, fig. 425. 
Plaxiphora Carmichaelis, Gray, P.Z.S. Lond. (1846), p. 68. 
,, ,, Haddon, Challenger Rep., p. 32. 

» >, H. and A. Adams, Gen. Rec. Moll., i. p. 481, and iii., t. 55, fig. 3. 

Chxtopleura Savatieri, Rochebrune, Bull. Soc. Phil. Paris (1880-81), p. 119; Miss. Sci. du 
Cap Horn. 
,, frigida, Rochebrune, I.e., p. 137, t. 91, figs. 5a, 5b. 



Bab.— Gough Island, April 22, 1904. Station 461, 
Scotia Bay, South Orkneys, rarely. Station 325. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 343 

Chsetopleura brucei, Iredale, sp. n. (Plate, figs. 24, 24a— c?.) 

Shell of medium size, ovate, depressed, girdle fleshy, densely covered with hairs, the 
longer being very prominent on a bed of shorter ones, appearing to be grouped and 
longest near the sutures. Valves broad, with a well-marked keel, though not very high, 
the posterior valve having the mucro about the anterior third. 

Shell smooth, the lateral areas being indicated by a very faintly raised ridge. 

Down the median keel of the five centre valves is a row of pustules which do not 
reach to the mucro, and two parallel rows can be seen on either side, these rows show- 
ing on the anterior portion of the posterior valve ; but on the first median valve this 
arrangement is not so apparent. 

Scattered radiating rows of similar pustules are seen on the anterior valve, where 
faint ridges are indicated ; similar sculpture is seen on the posterior part of the end 
valve. On the pleural areas of the median valves scattered pustules are also present, 
whilst the lateral areas have them also few and scattered. Otherwise the only feature 
is the concentric growth-ridges, which are well marked on each ridge, indicating regular 
growth in still water. 

The internal features are, as noted by Pilsbry for C. peruviana, Lamk. (Man. 
Conch., xiv. p. 29, 1892), the anterior valve with 9, central valves 1, and posterior 
valve 9 slits. 

Hab. — Scotia Bay, South Orkneys. Station 325. One fine specimen only. 
Agrees closely with C. peruviana, Lamk., and seems to be the first record for the 
genus from east of South America. (T. Iredale.) 

Lepidopleurus pagenstecheri, Pfeffer. 

Leptvchiton pagenstecheri, Pfeffer, Jahrb. hamburg. wissenschaftlichen Anstalten, iii 
Jahrgang, p. 107, t. iii. fig. 3 (1886). 

Hab. — Scotia Bay, 9-10 fathoms. Station 325. 

Thiele considers this Chiton conspecific with L. Jcerguelensis, Haddon, from 
Kerguelen Island, but Iredale does not accept this conclusion, though admitting the 
close alliance of the two species. 

Hemiarthrum setulosum, Carpenter. 

Hemiarthrum setulosum, Carpenter, MS., p. 13. — Dall, Bull. U.S. Nat. Mus. ii., (1876), p. 44. 
— Haddon, "Challenger" Polyplacophora, p. 14, t. i. fig. 4; t. ii. fig. 4a, 1. — Martens 
and Pfeffer, Jahrb. des hamburg. wissenschaftlichen Anstalten, iii. p. 108, t. iii. fig. 4 
(1886). 

Hab. — Station 325, Scotia Bay, 9-10 fathoms, on Fuci and other Algse. 

Very small and juvenile specimens, probably referable to the above. Iredale also 
doubts the identity of the South Georgian Hemiarthrum with that described by Dall 
from Kerguelen Land, but more material is wanted for comparison. 



344 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Order Prosobranchiata. 
Sub-order diotocardia. 

Section Zygobranchiata. 

Family Fissurellidse. 

§ Emarginulidse. 

Puncturella noachina (L.). 

Patella noacliina, Linnaeus, Mantissa, p. 551. 

Puncturella noachina, Lowe, Zool. Journ., iii. p. 78 (1827). 

Cemoria princeps, Mighels, Proc. Boston Soc. Nat. Hist. (1841), p. 49. 

Rimula galeata, Gould, U.S. Explor. Exped., p. 369, t. xxxi. figs. 476, 477. 

Hab. — Trawl, Burdwood Bank, Station 346, south of the Falkland Islands, lat. 
54° 25' S., long. 57° 32' W., December 1, 1903. 

Bleached but perfect specimens of a British and North European species, also 
known to extend to the Falkland Islands and Straits of Magellan. It is likewise 
recorded by Dr Hermann Strebel,* from Berkeley Sound, lat. 51° 53' S., long. 58° W. 

We include under the name noachina (L.) various forms, e.g. conica, D'Orb., 
falklandiana, A. Ad., cognata, Gould, and galeata, Gould. It is most probable that 
the gatherings from Burdwood Bank would come under the name mentioned second, 
falklandiana. 

Family Pleurotomariidw. 

Scissurella eucharista, sp. n. (Plate, figs. 1, la). 

S. testa perminuta, globulosa, delicatissima, alba, naticoide, paullum elevata, anfractibus 4, quorum 
apicalis fere' immersus, penultimo inflato, tumescente, ultimo epidermide evanida pallide straminea contecta, 
infra suturam leniter planato, deinde bicarinato, quarum inter fines scissura extensa, angusta, caetera super- 
ficie delicate sub lente spiraliter tenuissime striata usque ad basim supra carinam radiatim leniter plicata, 
umbilico fere' clauso, apertura rotunda, labro rotundo, tenuissimo. 

Alt. 1, diam. -75 mm. 

Hab. — Burdwood Bank, 56 fathoms, trawled. Station 346. 

A perfect example of one of the smallest shells possible, and yet full of character. 
We have compared it with the majority of the genus, and find it stands out con- 
spicuously in general roundness of outline, the double carination, within which, towards 
the aperture, is situate the narrow extended slit, not causing, as is usual, an angular 
appearance. Indeed, in form it is almost naticoid. Below the carinas, the surface to 
the base is transversely very finely striate, the umbilicus appears partly covered, the 
outer lip is round and extremely thin. Somewhat of the same form as Sc. conica, D'Orb., 
also from Southern waters ; but in our species the slit is situate much nearer the suture, 
that of conica being almost median. (ev^apio-ro?, elegant, agreeable.) 

* Schwed. Sudpolar Exped. (1908), p. 79. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 345 

Scissurella euglypta, Pels. 

Scissurella euglypta, P. Pelseneer, Voy. du S.T. " Belgica" : Zoologie, p. 17, pi. iv. 
figs. 43-45 (1903). 

Hab. — Trawl, Burdwood Bank, at 56 fathoms. Station 346. 

Only one imperfect specimen, but recognisable. 

Scissurella supraplicata, Sm. 

Scissurella supraplicata, E. A. Smith, Ann. and Mag. N.H., xvi. p. 72 (1875). 

„ ,, ,, Phil. Trans. Roy. Soc. London, vol. clxviii. p. 176, 

pi. ix. figs. 5, 5a (1879). 

Hab. — Trawl, Burdwood Bank, at 56 fathoms. Station 346. 

Several examples, mostly imperfect, of this pretty species, striking on account of 

its very marked plication above the double keel. 

Scissurella timora, sp. n. (Plate, figs. 2, 2a). 

Sc. testa minuta, tenuissima, alba, epidermide straminea omnino contecta, depresso-effusa, anfractibus 4" 
apicalibus parvis, ultimo lato, supra ad peripheriam planato, radiatim lineis obliquis tenuibus prsedito, 
scissura angusta, profunda, infra ad basim concentrice trilirato, apertura ovata, intus alba, labro pertenui, 
columella paullum incrassata, fere^ recta. 

Alt. 1, diam. 175 mm. 

Ha b. —Station 325, Scotia Bay, South Orkneys, 9-10 fathoms, on Macrocystis 
pyrifera and other large Fuci. 

A depressed, obliquely effuse little species, of which but few examples occurred, all 
in live condition, covered with straw-coloured epidermis. The upper part of the body 
whorl is not so conspicuously radiate as in many species ; the slit is narrow, deep, its 
edges being carinate. (-n^w^o'?, honoured.) 

Section Azygobranchiata. 

Family Gyclostrematidte. 

Cyclostrema calypso, sp. n. (Plate, fig. 3). 

C. testa perminuta, anguste sed profundi umbilicata, conica, alba, delicatula, anfractibus ad 5, inclusis 
apicalibus duobus lsevibus, cseteris arete longitudinaliter liratis, et spiraliter decussatim striatulis, numero 
lirarum ultimi anfractus ad quadraginta, anfractibus omnibus ventricosis, ad suturas multum impressis, 
apertura rotunda, peristomate continue 

Alt. 1, diam. 1*15 mm. 

Hab.— Trawl, Burdwood Bank, lat. 54° 25' S., long. 57° 32' W., at 56 fathoms. 
Station 346. 

Exceedingly minute, resembling C. decussatum, Pelseneer,* in many ways, but 
differing in (a) size, and (b) in fine and close longitudinal liration. To C. conicum, 
Watson, collected during the Challenger Expedition (Station 24), it likewise is akin ; 
but in this species, more than double the dimensions to begin with, the lamellae are 
much stronger proportionately, and fewer in number than in either C. decussatum 
or C. calypso. 

* P. Pelseneer, Voy. du S. Y. " Belgica," p. 19, pi. v. fig. 48 (1903). 



346 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Cyclostrema coatsianum, sp. n. (Plate, figs. 4, 4a). 

C. testa parva, alba, solidula, profunde umbilicata, elegantissime sculpta, anfractibus 4, quorum duo 
apicales nitidi, albi, la3ves, duobus caeteris longitudinaliter Eequicostatis, costis lsevibus, incrassatis, subflexuosis, 
penultimo supra planato, ultimo spiraliter quadricarinato, carina obtusa infra suturas, binis ad peripheriam, 
pra;ditis, simul ac ad basim, interstitiis omnibus subquadratis et fenestratis, regione umbilicari profunda, 
verticali, apertura rotunda, peristomate crassiusculo, continuo. 

Alt. 1, diam. 2'25 mm. 

Hob.— Trawl, Burdwood Bank, lat. 54° 25' S., long. 57° 32' W., 56 fathoms. 
Station 346. 

A very small, solid, white, boldly but elegantly sculptured Cyclostrema, the 
nearest ally being C. micans, A. Ad., from the eastern tropics, known in Indian seas 
as Liotia pulchella* Dunker. This species is somewhat larger, and the pattern of 
sculpture is different. We name this species in honour of Mr James Coats, of Ferguslie 
House, Paisley, through whose generosity the Scottish National Antarctic Expedition 
was equipped with funds, and whose regretted death, by a strange coincidence, occurred 
just after this description had been drawn up, on March 22, 1912. 

Cyclostrema gaudens, sp. n. (Plate, figs. 5, 5a, 5b). 

C. testa minutissima, profunde umbilicata, depresso-discoidali, supra planiuscula, alba, anfractibus ad 
3|, quorum apex ipse depressus, perlsevis, ultimo ad peripheriam obtuse carinato, undique longitudinaliter arete 
lirato, liris circa viginti-duabus, apud basim circa umbilicum obscure spiraliter carinato, apertura rotunda, 
peristomate tenui, fere continuo, operculo corneo, multispirali, nucleo centrali. 

Alt. "75, diam. 1 mm. 

Hob. — Station 346, trawl, 56 fathoms, Burdwood Bank. 

Slightly allied to the preceding, but much differing in sculpture, especially in the 
suppression of the prominent peripheral keeling of the body whorl. Judging from 
the figure, there is an affinity to C. alveolatum, Jouss.,+ described from an unknown 
locality, the dimensions being only slightly less ; the interstices, however, between the 
tfexuous costse do not appear, in our species, to be spirally striate, as is the case with 
Jousseaume's species. 

Cyclostrema meridionals, sp. n. (Plate, figs. 6, 6a, 22, 22a). 

C. testa minutissima, depresso-trochoide, delicata, tenui, pallide albo-cinerea, epidermide fugitiva 
straminea omnino contecta, profunde umbilicata, anfractibus 4, quorum duo apicales tumescentes, albi, 
perlseves, ceteris duobus — penultimo uni-, ultimo anfractu spiraliter bicarinato, apertura rotunda, peristomate 
continuo, paullulum incrassato, apud basim circa umbilicum crenello-carinato, operculo multispirali, corneo, 
nucleo fere centrali. 

Alt. -75, diam. - 50 mm. 

Hob. — Gregariously, on various Algae (Fucus and Macrocystis), Station 325, Scotia 
Bay, South Orkneys, 9-10 fathoms. 

This well-defined but very minute species is evidently the same as that recorded 
from the same islands by Dr E. Lamy,^ and considered a non -adult form of an unknown 

* A. Adams, P.Z.H. (1850), p. 44 ; Dunker, Mai. Blatt, vol. vi. p. 225 (1860). 
t Guerin, Mag., p. 392, pi. xix. fig. 4 (1872). 

| Moll. Rey. Arct. Norv., p. 135, pi. xxi. tig. 1 (1908); Bull. Mus. Nat. d'Hist. Naturelle (1906), Paris, p. 123, 
(1910) p. 323. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 347 

species of Margarita. To us, and several other malacologists who have examined it 
with care, it not only appears almost full-grown, but with some confidence is now pro- 
posed to be included in the genus Cyclostrema, at all events provisionally ; for this 
genus is somewhat multifarious already in its component parts, and much needs the 
services of a special monographer. 

The nuclear whorls are, it is true, slightly nepionic, and shapelessly turgid, but the 
penultimate and body whorls are very well sculptured and denned, being acutely spirally 
bicarinate. Around the umbilicus, likewise, a third keel, crenulate, and not so acute, 
revolves. A pale straw-coloured epidermis covers the whole surface uniformly. The 
operculum, for microscopic aid in the examination of which we are much indebted to 
Messrs E. A. Smith and Eobson of the British Museum (Natural History), is dark red- 
brown, with nucleus not quite central, and multispiral. This we take the opportunity 
also to figure (fig. 22a). 

Galliostoma modestulum, Strebel. 

Calliostoma modestulum, H. Strebel, Schived. Sudpolar Exped., p. 70, Taf. i. fig. 13 a, b (1908). 

Hah. — Station 346, Burdwood Bank, 56 fathoms, from Sponge. 

Two very young specimens, trochoid in form, with the upper whorls elegantly 
spirally lirate, we assign to the above name with a little doubt. The original type 
came from the West Falklands, lat. 52° 29' S., long. 60° 36' W., dredged at 197 metres 
(Strebel). 

With this occurred Photinula expansa, Sowb., and one broken example of a 
beautifully nacreous shell, which, judging from the figure,* may be Calliostoma 
mobiusi, Strebel. Our specimen is more trochoid than photinuloid, though it possesses 
some characters of the latter, and is lightly spirally grooved, these being most con- 
spicuous at the periphery of the body whorl. Dimensions : alt. 10, diam. 12 mm. It 
likewise may be compared with Photinula Craivshayi^ Sm., from Christmas Island, but 
the whorls are not ventricose. It is unfortunately somewhat broken ; the operculum is 
present, being horny and multispiral. 

Sub-order monotocardia. 
(a) Ptenoglossa. 
Family Scalidse. 

Scala magellanica, Phil. 
Scalaria magellanica, Philippi, Archivfiir Naturg., vol. i. p. 65 (1845). 
Hab. — Station 346, Burdwood Bank, 56 fathoms, in Sponge. 

Only very imperfect specimens, either very young or broken fragments ; enough 
however, to identify the species. 

* Strebel, Moll, der Magalhaen. Prov., ii. p. 133, Taf. v. fig. 22. 
t Smith, Proc. Malac. Soc. Lond., vi. p. 335, fig. 2. 



348 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

(6) Tsenioglossa. 

Family Naticidse. 

Natica impervia, Phil. 

Natica impervia, Philippi, Archiv filr Naturg., vol. i. p. 65 (1845). 

Hab. — Station 346, from Sponge at 56 fathoms. 
Only very dead and featureless specimens. 

Family Trichotropidse. 
Trichotropis antarctica, sp. n. (Plate, fig. 7). 

T. testa parva, imperforata, fragili, breviter fusiformi, sordide alba, anfractibus 6, quorum apicales tres 
detriti, sublseves, cseteris tenuiter et arete' decussatis, interstitiis quadratulis, ultimo cgeteros exsuperante, 
apertura ovata, labro multum expanses, inflato, margine columellari paullum excavate 

Alt. 5*75, diam. (oris) 3-50 mm. 

Hab. — Trawl, Burdwood Bank, at 56 fathoms. Station 346. 

A very interesting form, and we deem it worthy of description, albeit the only 
specimen is imperfect, and the outer lip infested with growth of a Bryozoon. It seems 
adult, and is comparable with no other member of the genus known to us. It is much 
smaller in all its parts than T. inornata, Hutton, from New Zealand. There is no sign 
of umbilication, and the epidermis is not present, being completely worn off. 

Calyptreea chinensis, L. 

Patella sinensis, Linnagus, Syst. Nat., ed. xii., p. 1257 (1769). 

G. sinensis, F. and H., ii. p. 463, pi. lx. figs. 3-5, and (animal), pi. B.B. figs. 8-13. 

Hab. — Burdwood Bank, south of the Falkland Islands. Station 346. 
Indistinguishable from the shell of northern climes, including Great Britain. 

Family Littorinidse. 

Littorina (Lsevilitorina) caliginosa (Gould). 

Hab. — An additional locality is Cape Pembroke, Falkland Islands, shore, 
February 2, 1904. Station 349. 

Littorina [Pellilitorina) pellita, v. Mts. 

Hab. — Additional locality for this species is Station 346, 56 fathoms, December 1, 
1903. Lat. 54° 25' S., long. 57° 32' W. Obtained from new species of Cephalodiscus, 
occasionally. 

Two more examples of Lacuna notorcadensis, M. and St., also occurred from the 
same locality as the type. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 349 

Lacuna abyssicola, sp. n. (Plate, figs. 8, 8a, 8b). 

L. testa parva, profundi umbilicata, albo-calcarea, epidermide brunnea contecta, fragili, vix solida, 
sinfractibus 5, quorum apex ipse mire immersus, duobus his proximis cum penultimis tumidulis, ultimo 
paullum effuso, lgevi, omnibus infra suturas canaliculatis et acute spiraliter carinatis, apud basim, circa 
umbilicum, crassi-carinato carinis binis, sulco interstitiali prsedito, apertura rotunda, labro tenui, margine 
columellari late reflexo. 

Alt. 2, lat. 2T5 mm. 

Hob.— Deposit No. 38, dredged March 18, 1904. Lat. 71° 22' S., long. 16° 34' W., 
1410 fathoms. Station 417. 

One specimen only, but adult and fairly perfect, save for a slight fracture of the 
outer lip. It appears nearly akin to L. nautiliformis , Jeffreys, or L. cincta of the same 
author, from the Atlantic, collected on the Porcupine Expedition, especially as 
regards the sculpture round the umbilical region, the thickened double carination with 
interstitial sulcus. Another feature of interest is the curiously immersed nucleus, 
and the strong canaliculation round the upper portion of each whorl, followed by an 
acute spiral keel. The substance of the shell is chalky white, covered with a dark- 
brown epidermis. The specific name proposed is given in consideration of the extreme 
depth at which it was dredged. 

Lacuna wandelensis, E. Lamy. 

Lacuna wandelensis, E. Lamy, Exped. Antarct. Francaise commandee par le Dr Jean Charcot : 
Moll., p. 5, pi. i. figs. 5, 6, 7 (Paris, 1906). 

Hab. — Station 325, Scotia Bay, South Orkneys, 9-10 fathoms, on Macrocystis and 
other large Fuci. 

A very few examples, and all in non-adult condition, belong almost certainly to 
this species. 

Family RissoicLr. 

Rissoa deserta, Sm. 

Rissoa deserta, E. A. Smith, Nat. Antarct. Exped. : Nat. Hist., vol. ii. p. 9, pi. ii. fig. 1 
(1907). 

Hab. — South Orkney Islands, Scotia Bay, 9-10 fathoms. Station 325. 
The specimens are dead, but seem to agree in form with the above species. 

Rissoa (Onoba) fiiostria, sp. n. (Plate, fig. 9). 

R. testa parva, paullum inflata, solidula, parum rimata, anfractibu-* i\, quorum apicales duo lseves, 
tumiduli, ceteris ventricosis, apud suturas impressis, arctissime spiraliter tenuiliratis, apertura ovata, alba, 
labro paullum effuso, baud multum incrassato, fere continuo. 

Long. 2, lat. L50 mm. 

Hab. — South Orkney Islands, Scotia Bay, 9-10 fathoms. Station 325. 

Allied to several Onobee, mostly described of recent years from deep-sea explora- 

TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 18). 53 



350 DR JAMES COSMO MELVILL AND ME ROBERT STANDEN ON THE 

tions, such as R. transenna* Wats., from Prince Edward Island, and R. aedonis^' of the 
same author, from Nightingale Island. R. gelida, E. A. Sm.,J is, perhaps, the nearest 
ally ; this is also an Antarctic species, and differs in the possession of an extra whorl, 
and being longer proportionately to its breadth, also in a lesser degree of ventricosity 
of whorl. Only two or three examples. 

Rissoa (Onoba) fuegoensis (Strebel). 

Rissoa (? Cingula) fuegoensis, H. Strebel, Schwed. Sudpolar Exped., p. 56, Taf. vi. fig. 90 
■ a, b (1908). 

Hah. — Burdwood Bank, Station 346, 56 fathoms. 

A straw-coloured, closely spirally lirate Rissoa, which we should consider as 
appertaining to the section Onoba in preference to Cingula. 

Rissoa (Onoba) paucilirata, sp. n. (Plate, fig. 10). 

R. testa ovata, anguste rimata, alba, epidermide tenuiter evanida straminea, interdum iridescente, 
contecta, anfractibus ad 5, ventricosulis, apud suturas multum impressis, quorum duo apicales nitidi, albi, 
Iseves, caeteris duobus fortiter spiral iter pauciliratis, liris penultimi duabus, ultimi anfractus septem vel 
octo, prseditis, apertura ovato-rotunda, peristomate tenui, margine columellari fere recto. 

Alt. 2-25, diam. 1-25 mm. 

Hah. — Burdwood Bank, Station 346, 56 fathoms. 

Conspicuous for its strong, spiral, carinated lirse, which are fewer in number than 
those possessed by its allies ; these spiral ridges seem much the same in the Aleutian 
species R. Aurivillii, Dall, § or R. brachia, Watson, || from Culebra Island, West Indies. 
This last, indeed, seems a very near ally, though quite distinct. 

Rissoa (Onoba) sulcata (Strebel). 

Rissoa (Cingula) sulcata, H. Strebel, Schwed. Sudpolar Exped., p. 56, Taf. vi. fig. 86 a, l, c (1908). 

Hob. — With the last species named, at 56 fathoms. Station 346. One specimen. 
The spiral sulci are interesting. In form it resembles R. paucilirata, but the 
essential characters are quite diverse. Colour inclined to reddish-fuscous. 

Rissoa (Ceratia) turqueti, E. Lamy. 
Rissoa (Ceratia) turqueti, E. Lamy, Exped. Antarct. Francaise Charcot, p. 6, pi. i. fig. 8 (1906). 
Hob. — With the preceding. One fine specimen in live condition, sub-pellucid, with 
faint relics of thin stramineous epidermis. Station 346. 

* Rep. Challenger Exped., xv. p. 620, pi. xlvi. fig. 10. t Ibid., p. 600, pi. xlv. fig. 5. 

\ Smith, Nat. Ant. Exped. : N.H., vol. ii. p. 9, pi. ii. fig. 5. § Proc. U.S. Nat. Mus., p. 309, pi. iv. fig. 8 (1 

|| Hep. Challenger Exped., xv. p. 599, pi. xlv. fig. 8. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 351 

Eatoniella kerguelenensis, Sm., forma major, Strebel. 

EatonielJa kerguelenensis, Smith, forma major, Hermann Strebel, Schwed. Sudpolar Exped., 
p. 57, Taf. iv. fig. 56 a-c (1908). 

Hah. — Station 325, Scotia Bay, South Orkneys, 9-10 fathoms. 

This larger form of a mollusc already reported by us, in our former paper, as occur- 
ring, in its typical condition, at Scotia Bay, South Orkneys, has likewise been discovered 
in some quantity in Bay A, of greater size and solidity, often encrusted with bryozoic 
and other growths. Colour very deep plumbeous. 

Family Cerithiidse. 
Cerithium pullum, Phil. ( = cselatum, Couthouy). 
Hob. — An additional locality is now given for this species, to that mentioned on 
p. 135 of our former paper, viz. Burdwood Bank, lat. 54° 25' S., long. 57° 32' W., in sponge. 
Station 346. 

We do not repeat the synonymy, which will be found at the page just quoted. 

Bittium brucei, sp. n. (Plate, fig. 11). 

B. testa minuta, solidula, cylindrica, castaneo-brunnea, anfractibus ad 8, apicalibus .... (?), cseteris apud 
suturas impressis, supernis bino, ultimo trino odine granulato regulariter prsedito, apud basim excavato, planato, 
apertura ovata, labro simplice, margine columellari crassiusculo. 

Long. 2*75, lat. 1 mm. 

Hob.— Dredge, Station 81, lat. 18° 24' S., long. 37° 58 W., 36 fathoms. 

A minute Cerithioid mollusc, which seems as if it should belong to the sub-genus 
Joculator, Hedley,* proposed for Cerithiopsis ridicula, Watson, and certain allies. At 
the same time it is so like Bittium minimum, T. Woods, well figured from a Tasmanian 
specimen by C. HEDLEY,f that it had better be included in that genus. 

Bittium burdwoodianum, sp. n. (Plate, fig. 12). 

C. testa fusiformi, brunneo-rufescente, parva, anfractibus ad 10, quorum apical es tres rufescentes, 
parum nitidi, lseves, vel simpliciter longitudinaliter costulati, cseteris ad suturas multum impressis, trino ordine 
gemmarum, ultimo quatuor ordinibus similibus, regulariter spiraliter prseditis, apertura ovata, labro paullum 
effuso, columella flexuosa. 

Alt. 4, diam. 1 mm. 

Hah. — From interior of Liothyrina. Station 346, Burdwood Bank, at 56 
fathoms, December 1, 1903. 

A little species, of simple character, inclined to a reddish hue, particularly as 
regards the apex and central portion of the various whorls, which are thrice spirally 
girt with regular rows of close grains, gemmulate and rounded. This might be 
considered a Cerithium by some authors. It is akin to B. bisculptum,\ Strebel, the 
apical whorls seemingly almost identical, and we consider these two species should 
stand in the same genus. 

* Proc. Linn. Soc. N.S. Wales (1909), p. 442. t Ibid. (1909), p. 722, fig. 20. 

% Schwed. Sudpolar Exped., p. 49, Taf. vi. fig. 92 a-b (1908). 



352 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Cerithiopsis macroura, sp. n. (Plate, fig. 13). 

C. testa elongato-fusiformi, parva, angusta, nitida, albo sfcraminea, anfractibus ad 10, quorum apicales duo 
vel tres nitidi, vitrei, perlaeves, bulbosi, cseteris paullum ventricosis, apud suturas impressis, undique longi- 
tudinaliter avctc costulatis, costulis anfractuum superiorum pro maxima parte lsevissimis, quatuor ultimis 
anfractibus spiraliter rugoso-liratis, liris ad juneturas costularum granulosis, apertura ovata, labro tenui, 
columella paullum producta, flexuosa, brevirostri. 

Alt. 3*55, diam. 1 mm. 

Hab. — Station 346, Burdwood Bank, 56 fathoms. 

A small species, but distinguished, as the specific name chosen would show, by 
its very attenuate, fusiform whorls, the last three or four swollen, caudate, shining, 
smoothly costulate, not spirally crossed with granose lirse, as are the lower whorls ; the 
columella is only slightly rostrate, the outer lip thin, the colour whitish straw. But 
few examples occurred. (fxaKpos oupa^ long-tailed.) 

Cerithiopsis malvinarum, M. and St. 

Cerithiopsis malvinarum (Strebel, MS.), Melvill and Standen, Trans. Roy. Soc. Edin., xlvi. 

pt. i., p. 135, pi. figs. 6, 6a. (1907). 
,, ,, Strebel, Schwed. Sudpolar Exped., Band vi., 1, p. 49, Taf. i. fig. 10 

a-c (1908). 

Hab. — Shore, Hearnden Water, Falkland Islands. Station 349. 

As mentioned in our first paper, we issued a description of this species in 1907, 
using, at Dr H. Strebel's request, the name he had given it in manuscript. The 
following year it was redescribed by him as " sp. nov.," and we are of opinion that he 
had not at that time seen our paper. The same remarks would apply also to Trophon 
minutus, M. and S. 



Family Turritellidse. 

Turritella algida, sp. n. (Plate, fig. 14). 

T. testa parva, attenuato-fusiformi, alba vel pallide straminea, solidula, anfractibus ad 9-10, ad suturas 
multum impressis, quorum apex ipse bulbosus, albus, lsevis, vitreus, huic proximus anfractus simili modo 
tumidus, lsevis, ceteris ad medium unicarinatis, carinis acutis, prominulis, antepenultimo et penultimo lira 
alia minore infra medium praeditis, ultimo inter carinam majorem et basim trilirato, apertura ovata, labro 
tenui. 

Long. 6, lat. 2 mm. 

Hab. — Trawl, Burdwood Bank, south of the Falklands, at 56 fathoms. Station 346. 

Very small, but apparently quite adult. Conspicuous for a distinct and prominent 
median keel, the three last whorls also being provided with a minor spiral lira below, 
and the body whorl, between the strong median keel and the base, possessing three such 
spiral lirations. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 353 

Mathilda rhigomaches, sp. n. (Plate, fig. 15). 

M. testa minuta, imperforata, fusiformi, delicata, pallide fuscescente, anfractibus 6^, quorum apicales 
2J heterostrophi, albi, lseves, bulbosi, cseteris apud suturas impressis, pulchre spiraliter carinatis, carinulis 
tribus anfractnum superiorum, ultimo quatuor, arctissiine lirulis longitudinalibus decoratis, interstitiis 
quadratis. apertura rotunda, labro tenui, margine columellari paullum excavato. 

Long. 2, lat. 1 mm. 

Hab. — Trawl, Burdwood Bank, at 56 fathoms. Station 346. 

In sculpture this little species resembles a Lovenella, especially L. austrina, 
Hedley, # from the opposite shores of Antarctica. It is only about a quarter of the 
size, however, of this shell, while the apex is heterostrophe, the peristome continuous. 
Fischer {Man. de Conch., p. 172, 1887) gives a list of Magellanic Mollusca, and 
includes a "Mathilda magellanica." This is evidently a " nomen nudum." No 
description can be found, and the name rests on no authority. The remarks of 
M. de Boury t will probably, in connection with this, be found of interest. (piyofAd^s, 
contending with cold.) 

(c) G-ymnoglossa. 

Family Eulimidse. 
Eulima antarctica, Strebel. 

Eulima antarctica, H. Strebel, Schwed. Sudpolar Exped., Band vi., 1, p. 65, Taf. vi. fig. 91 a-c 
(1908). 

Hab. — Trawl, Burdwood Bank, south of the Falkland Islands, 56 fathoms. 
Station 346. 

One specimen, live, but hardly full-grown. 

Family Pyramidellidse. 
Turbonilla smithii, Pfeffer. 

Turbonilla smithii, G. Pfeffer, MS. in H. Strebel, Mollush. der Magalhaen. Prov., p. 659, Taf. 
xxiii. fig. 42 a-d (1905). 

Hab. — Trawl, Burdwood Bank, at 56 fathoms. Station 346. 

One example, immature, but with sufficient characters to pronounce fairly certainly. 

Turbonilla xenophyes, sp. n. (Plate, figs. 16, 16a). 

T. testa aciculato-fusiformi, delicata, subpellucida, albo-lactea vel pallide straminea, paullum nitente, 
anfractibus 9, quorum apicales bulbosi, tumidi, leniter heterostrophi, caeteris ventricosulis, ad suturas impressis, 
sub lente delieatissime longitudinaliter liratis, in speciminibus quibusdam fer^ vel omnino lsevibus, apertura 
ovata, peristomate tenui, columella simplice. 

Long. 2-75, lat. -75 mm. 

* Report Brit. Antarct. Exped., 1907-9 (Shackleton), vol. ii., part i. p. 5 (pi. i. fig. 7) (1911). 
t Journ. de Conch., vol. xxxi. p. 118 (1883). 



354 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Hab. — Trawl, Burdwood Bank, south of the Falkland Islands, 56 fathoms. 
Station 346. 

A curious species, and one of which we are not quite sure of the proper position 
generically. It seems, however, to agree with Turbonilla in more than one feature. 
It is very delicate, resembling a terrestrial Opeas or others of the family Stenogyridse, 
both in substance and form. Several examples occurred, the live specimens retaining 
a subpellucid appearance and dull straw-colour, (^evocpvrw, strange of form.) 

(d) Rachiglossa. 

Family Muricidse. 

Trophon falklandicus, Strebel. 

Trophon falklandicus, H. Strebel, Schwed. Sudpolar Exped., Band vi., 1, p. 39, Taf. i. fig. 8 a-c (1908). 

Hab. — Burdwood Bank, at 56 fathoms. Station 346. 

Very young specimens are probably referable to this species. Another, judging 
alone from the plate (Strebel, Zool. Jahrbuch, Band xxi., Taf. vii. fig. 56, 1904), 
might belong to T. Paessleri, Streb. We cannot, however, help feeling that too many 
species have been created in such a variable assemblage as this section of the genus 
Trophon presents. 

Trophon minutus, M. and St. 

Trophon minutus (Strebel, MS.), Melvill and Stauden, Trans. Roy. Soc. Edin., xlvi., pt. i., 
p. 137, pi. figs. 7, la (1907). 
„ ,, Strebel, Schwed. Sudpolar Exped., Band vi., 1, p. 44, Taf. iv. fig. 47 

a, b (1908). 

Hab. — An additional locality to that mentioned in our former paper is Scotia Bay, 
South Orkneys, at 9-10 fathoms. Station 325. 

Three or four more examples occurred, but the species is evidently rare. For the 
nomenclature of this species, and its authorship, see remarks under Cerithiopsis 
malvinarum. 

Trophon philippianus, Dkr. 

Hab. — Also from Burdwood Bank, at 56 fathoms, all the specimens being in very 
young condition, and found in Sponge. Station 346. 

Antistreptus rnagellanicus, Dall. 

Antistreptus rnagellanicus, W. H. Dall, Proc. U.S. Nat. Mus., xxiv. p. 532 (1902). 

„ ,, Dall, Bull. Mus. Comp. Zool. Harvard, vol. xliii. p. 315, pi. xv. 

fig. 14(1905). 
Glypteuthria contraria, H. Strebel, Schived. Sudpolar Exped., Band vi., 1, p. 29, pi. i. 

figs. 4 a-c (1908). 

Hab. — Burdwood Bank, Station 346, at 56 fathoms. 
Two examples of this small, but curious, sinistral species. 



MARINE MOLLUSC A OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 355 

Family Buccinidse. 
Chrysodomus (Sipho) crassicostatus, M. and St. 

Chrysodomus (Sipho) crassicostatus, Melvill and Standen, Trans. Roy. Soc. Edin., vol. xlvi. 

part i., p. 138, pi. figs. 10, 10a (1907). 
Sipho (IMohnia) astrolabiensis, H. Strebel, Schwed. Sudpolar Exped., Band vi., 1, p. 31, 

Taf. iii. fig. 37 a-d (1908). 

One specimen of Sipho astrolabiensis occurred in lat. 63° 9' S., long. 58° 17' W., 
at Astrolabe Island. 

From the figure, there can be no doubt of its identity with our *S'. crassicostatus, 
described the year previously (1907). More examples came to hand from the locality 
already given by us, viz. Scotia Bay, South Orkneys, at 9-10 fathoms, Station 325; 
and we have now seen it likewise from Burdwood Bank, at 56 fathoms, Station 346. 

Euthria rosea, Homb. and Jacq. 

Euthria rosea, Hombron et Jacquinot, Voyage au Pole Sud, v. p. 107, tab. xxv. figs. 4, 5. 
,, ,, Strebel, Mollusk. der Magalhaen. Prov., p. 616, Taf. xxi. figs. 1-4 (1905). 

Hab. — Burdwood Bank, from Sponge, at 56 fathoms. Station 346. 

Family Mitridse. 
Mitra (Volutomitra) porcellana, sp. n. (Plate, fig. 21). 

M. (V.) testa eleganter fusiformi, nitidissima, Candida, porcellana, anfractibus ad 6 (?), apicalibus . . . ? 
cseteris nequaquam suturaliter impr-essis, politissimis, ultimo prolongato, apertura anguste oblonga, labro 
tenui, columella obliquante, quadriplicata, plicis obliquis. 

Long. 14, lat. 6 mm. (sp. imperfecta). 

Hab. — Scotia Bay, South Orkneys, 9-10 fathoms, Station 325 ; also trawl, Burdwood 
Bank, 56 fathoms, Station 346. 

Only two examples of this beautiful, polished white porcellanous shell have as 
yet occurred, one from each locality, widely differing from other Volutomitrse known 
to us ; its narrow aperture, obliquely quadriplicate columella, are distinguishing 
characteristics. Very unfortunately, in neither specimen, owing to breakage, do the 
apical whorls appear, so several points remain for the present a matter of conjecture. 

(e) Toxoglossa. 

Family Conidse. 

Beta anderssoni, Strebel. 
Beta anderssoni, H. Strebel, Schwed. Sudpolar Exped., p. 14, Taf. ii. fig. 24 a-d (1908). 

Hab.— Station 346, at 56 fathoms, December 1, 1903. 



356 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Judging from figure and description quoted above, this interesting Beta is identical 
with specimens found at Seymour Island, Grahamland, by the Swedish expedition. 



Bela fulvicans, Strebel. 

Bela fulvicans, H. Strebel, Sclnoed. Sudpolar Exped., p. 15, Taf. ii. fig. 25 a-d (1908). 

Hab. — Burdwood Bank, from Sponge, at 56 fathoms. Station 346. 
An imperfect, bleached specimen seems, from the sculpture, to be the above 
species, which occurred both in South Georgia Islands and in Grahamland. 

? Thesbia sp. 

Hob. — Burdwood Bank, from Sponge, at 56 fathoms. Station 346. 

One example, more imperfect than the preceding, of a bleached shell, showing 
faint flexuous oblique longitudinal costellation, mouth narrow oblong, whorls fairly 
smooth, hardly impressed at the sutures. Dimensions : long. 13, lat. 5 mm. It 
is quite impossible to differentiate it further. 

Savatieria concinna, sp. n. (Plate, fig. 17). 

S. testa ovato-fusiformi, compacta, solidula, subpellucente, albida, anfractibus 6, quorum apicales 
duo bulbosi, vitrei, nitidi, perlseves, caeteris a pud suturas impressis, subventricosis, longitudinaliter arete 
costulatis, costis crassiusculis, gemmatis, ultimo anfractu infra medium evanidis, deinde ad basim spiraliter 
sulculoso, numero costularum anfractus ultimi circa 22, apertura ovata, labro simplice, columella paruni 
incrassata, canali vix prolongata. 

Long. 4-55, lat. 2 mm. 

Hab. — Trawl, Burdwood Bank, Station 346, 56 fathoms, December 1, 1903. 

Savatieria is a small genus, peculiar to these regions, diagnosed by B,ochebrtjne 
and Mabille. It is nearly allied to Bela, differing principally in the abbreviated 
canal, whorls peculiarly impressed suturally, and more distinct elaboration of 
sculpture. Several species have lately been published by Dr Hermann Strebel, 
and to one of them, S. molinzv, our species is akin, differing mainly in sculpture, 
being supplied with nearly double the number of longitudinal ribs, while the 
gemmate beading is more pronounced in S. concinna. Only one example, happily 
in first-class condition at the time of description, was procured, though unfortunately 
it was accidentally broken at the mouth before it could be figured. We consider that 
Lachesis meridionalis, E. A. Sm., # is synonymic with Savatieria molinse, Strebel, 1905, 
and has priority of twenty-four years over it. 

* J'roc. Zool. Soc. Lond., 1881, p. 28, pi. iv. fig. 3. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 357 

Family Cancellariidse. 

Paradmete typica, Strebel. 

Paradmete typica, H. Strebel, Schwed. Sudpolar Exped., Band vi., 1, p. 22, Taf. iii. fig. 35 a-f (1908). 

Hab. — Burdwood Bank, Station 346, at 56 fathoms, December 1, 1903. 
Thus showing a considerable extension in range. Only one specimen, but in 
good condition. 

Admete limneasformis, Sm. 

1 Admete limnemformis, E. A. Smith, Phil. Trans. Roy. Soc. Lond., clxviii. p. 172, pi. ix. 
fig. 4 (1879). 

Hab. — Trawl, Burdwood Bank, at 56 fathoms. Station 346. 

One example, in good condition, exactly agreeing with the type, from Kerguelen 
Land. We should hardly be prepared to suggest placing this in Dr Strebel's new 
genus Paradmete. Mr Charles Hedley has lately hinted at its possible reception 
into the genus Odostomiopsis, Thiele, and this is well worthy of consideration. The 
shell is small, white, semi-transparent, and, as the trivial name, so well chosen, 
suggests, almost an exact reproduction of Limnsea peregra, Mull., in miniature. 

Order Opisthobranchiata. 

Sub-order tectibranchiata. 

(a) Bulloidea. 

Family Tornatinidse. 

Retusa antarctica, sp. n. (Plate, fig. 20). 

R. testa delicata, parva, ovato-fusiformi, rimata, pallidissime livido-virescente, perlsevi, subpelhicida, 
anfractibus 4, quorum apicales duo tumescentes, cseteris ad suturas rotunde gradatim impressis, ultimo 
magno, laevi, apertura ovata, labro sinuato, vix erassiusculo, columella obliqua. 

Alt. 3-25, diam. 1-75 mm. 

Hab. — Scotia Bay, South Orkneys, 9-10 fathoms. Station 325. 
A small, plain, greenish-livid species, translucent, very smooth, with swollen nuclear 
whorls, and roundly shouldered at the sutures. 

Retusa truncatula (Brug.). 

This widely distributed species, the full synonymy of which we gave in our last 
paper (loc. cit., p. 141), and which is hardly distinguishable from the British form, 
also occurred at the Burdwood Bank locality, Station 346, 56 fathoms. 

Fragments of others of this order, belonging to the genera Cylichna and Philine, were 
TRANS. ROY. SOC. EDIN., VOL. XLVIII. PART II. (NO. 18). 54 



358 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

dredged either from the same or neighbouring localities, but none in a condition to 
examine seriously. 

Sub-order pteropoda. 

Section Thecosomata, de Blainville. 

Cavolinia tridentata (Forskal). 

1773. Anomia tridentata, Forskal, Descriptiones animalium quse in itinere orientali observavit, 

p. 124. 
1791. Cavolinia natans, Abildgaard, "Nyere Efterretning om det Skaldyr som Forskal har 

beskrevet under Navnet Anomia tridentata," Sh-iv. naturhist. Selsk., Bd. i., Heft 2, 

pi. x. 
1801. Hyalsea cornea, Lamarck, Systhne des animaux sans vertebres, p. 140. 
1804. Hyalsea papilionacea, Bary de St Vincent, Voyage dans les quatre principles ties des 

mers d'Afrique, t. i. p. 137, pi. v. fig. 1. 
1810. Hyale teniobranche, Peron et Lesueur, " Histoire de la famille des Mollusques 

Pteropodes," Ann. Mus. Hist. Nat. Paris, t. xv., pi. ii. fig. 13. 
1813. Hyalsea peroni, Lesueur, "Memoire sur quelques animaux mollusques, etc.," Nouv. 

Bull. Soc. Pldlom., t. iii. p. 284. 
1813. Hyalsea chemnitziana, Lesueur, ibid., p. 284. 
1816. Hyalsea austrulis, Peron, Voyage de deeouvertes aux terres australes, t. i , pi. xxxi. fig. 5 

(sine descriptione). 
1821. Hyalsea forskahlii, Lesueur, MS., in de Blainville, "Hyale," Diet. d. Sci. Nat., t. xxii. 

p. 79. 
1836. Hyalsea affinis, d'Orbigny, Voyage dans VAmerique meridionale, t. v. p. 91, pi. v. 

figs. 6-10. 
1848. Hyalsea truncata, Krauss, Sud-africanisc/ie Mollusken, p. 34, pi. ii. fig. 12 (non 

Lesueur). 
1859. Cavolinia telemus, A. Adams, "On the Synonyms and Habits of Cavolinia, Diacria, and 

Pleuropus," Ann. and Mag. Nat. Hist, ser. 3, t. iii. p. 44. 
1877. Hyalsea cumingii, Sowerby, in Reeve, Conchologia iconica, t. xx., Pteropoda, fig. 5. 

Hab.—L&t. 39° 58' S., long. 8° 36' W., tow-net, surface, temp. 55°'2. 
Many living specimens, large and fine. Between Stations 470 and 471. 



Class SCAPHOPODA. 

Dentalium eupatrides, M. and St. 

Dentalium eupatrides, Melvill and Standen, Trans. Roy. Soc. Edin., vol. xlvi., part i., p. 142, 
pi. fig. 12 (1907). 

Hub. — The original locality of this fine smooth abyssal species was accidentally 
omitted in our first paper. It occurred, with the other species chronicled, 
I), megathyris, Dall, in lat. 71° 22' S., long. 16° 34' W., at 1410 fathoms, 
Station 417. Many fragmentary portions of probably the same shells have been 
dredged up from Station 420, at 2620 fathoms. 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 359 

Dentaliwn megathyris, Dall. 

Dentalium megathyris, Dall, Proc. U.S. Nat. Mus., xii. p. 293, pi. ix. fig. 1 (1889). 
„ „ Stearns, Proc. U.S. Nat. Mus., xvi. p. 424 (1893). 

Hab.—L&t. 7\° 22' S., long. 16° 34' W., 1410 fathoms. Station 417. 

In our first report we allocated a large Dentalium dredged from 1410 fathoms to 
D. Shoplandi, Jouss., as it agreed with specimens so named in the British Museum 
from " near Aden." We have since received from the same station a large fragment 
of the upper part of a living specimen, evidently snapped off by the dredge, and a 
number of smaller fragments. Critical examination of these has led us to conclude 
that our specimens are identical with D. megathyris, Dall, which has occurred off 
Chiloe Island and south-east Chili in 1050 and 1342 fathoms, in the Gulf of Panama 
in 2282 fathoms, and other localities in the Panamic region. It is significant that in 
company with this Dentalium, both in the Gulf of Panama and in the 1410 fathoms 
locality, the Brachiopod, Macandrevia diamantina, Dall, should also occur. The 
descriptions and figures of D. megathyris and D. Shoplandi, as given by Tryon, are so 
widely different in every respect, both as to dimensions and sculpture, and other minor 
details, that although our specimens agree so well with the British Museum examples 
purporting to come from Aden, we now are inclined to refer them to D. megathyris, 
as, even if this species should ultimately be proved to be an extreme form of 
D. Shoplandi, that specific name would become a synonym — D. megathyris, Dall, 
having priority of five years. From a careful study of the material and literature at 
our command we cannot help thinking that D. megathyris, Dall, D. Shoplandi, Jouss., 
D. ceras, Watson, and perhaps D. majorinum, Eocheb. and Mab., may eventually 
prove to be but forms of one variable gigantic longitudinally costate Dentalium in 
the southern hemisphere, radiating towards the Atlantic as well as the Pacific Ocean, 
and inhabiting everywhere very deep water, where the great pressure, darkness, and 
equable temperature render it possible for it to range through many degrees of latitude. 



Class PELECYPODA. 

Order Protobranchiata. 

Family Nuculidse. 

Yoldia profundorum, sp. n. (Plate, figs. 18, 18a, 186). 

Y. testa parva, tumida, nitida, lsevi, insequilaterali, periostraco plumbeo-olivaceo contecta, antice 
rotimdata, postice paullulum producto, umbonibus erosis, approximatis, haud prominulis, ligamento obscuro, 
lineari, cardinibus utriusque valvse decern denticulis utrinque prseditis, pagina interna nitida, albo-lactea, 
sinu palliali parvo. 

Alt. 3, lat. 4-50 mm. 



360 DR JAMES COSMO MELVTLL AND MR ROBERT STANDEN ON THE 

Hab.— Deposit No. 38, dredged March 18, 1904, lat. 71° 22' S., long. 16° 34' W., 
1410 fathoms. Station 417. 

A small, tumid, smoothish, slightly inequilateral Yoldia, the anterior side 
rounded, the posterior somewhat produced, to which Y. (Sarepta) abyssicola, Smith,* 
from Station 246, Challenger Expedition, Mid North Pacific, at 2050 fathoms, and also 
Station 281, Mid South Pacific, at 2385 fathoms, seems somewhat allied. That species, 
however, appears more distinctly abbreviate posteriorly, and higher in proportion 
to its width. Y. ecaudata, Pelseneer,t may likewise be compared, a species which 
is closely akin to Y. abyssicola. This was obtained during the voyage of the 
Belgica in the Antarctic region, at a depth of 400-500 metres. Again, Y. Valettei, 
Lamy, from the South Orkneys, where an example was found in the stomach of a penguin, 
is much of the same outward form, but less than half the dimensions (2'2 x 1 "65 x 1*5 mm.), 
and the teeth are only six in number on either side. The epidermis is likewise named 
as " flava " in contradistinction to " plumbea " or " olivacea." 

Nucula pisum, Sowb. 
Nucula pisum, Sowerby, Thes. Conch., iii. p. 153, pi. ccxxix. fig. 133. 
Hob. — Falkland Islands, local, but gregarious. Station 118. 

Order Filibranchiata. 

Sub-order arcacea. 

Family Arcidse. 

Area (Bathyarca) strebeli, M. and St. 

Area (Bathyarca) strebeli, Melvill and Standen, Trans. Roy. Soc. Edin., vol. xlvi., 
part i., p. 144, pi. figs. 13, 13a (1907). 

Hob. — Two additional localities can be now given, as follows : — 
Station 420. Dredged at 2620 fathoms. One specimen. 

„ 291. Lat. 67° 33' S., long. 36° 35' W., 2500 fathoms, March 7, L903. 

Limopsis longipilosa, Pels. 
Limopsis longipilosa, P. Pelseneer, Voy. du S.Y. "Belgica" : Zoologie, p. 25, figs. 89, 90 (1903). 

Hob.— Dredged in lat. 71° 22' S., long. 16° 34' W., at 1410 fathoms, March 18, 
1904. Station 417. 

One fairly perfect specimen, probably referable to the above. 

[Very imperfect examples of another Limopsis, solid, small, equilateral, covered 
with thin, short-bristled epidermis, also occurred at Burdwood Bank, 50 fathoms.] 

* Rep. Challenger Expedition, " Lamellibranchia," pi. xx. figs. 6, 6a, 66. 

t Voy. du S.Y. "Belgica" : Zoologie, par Paul Pelseneek, p. 22, figs. 77, 78 (1903). 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 361 

Sub-order mytilacea. 

Family Mytilidse. 
Philobrya sublsevis, Pels. 
Philobrya sublsevis, P. Pelseneer, Voy. du S.Y. " Belgica" : Zoologie, p. 25, figs. 93, 94 (1903). 
Hob.— Station 346, lat. 54° 25' S., long. 57° 32' W., at 56 fathoms, January 1, 
1903. 

Philobrya wandelensis, Lamy. 

Philohryu wandelensis, E. Lamy, fiaped. Antarct. Francaise Charcot, 1903-5, p. 16, pi. i. 
figs. 15, 16 (1906). 

Hob. — Trawl, Burclwood Bank, 56 fathoms. Station 346. 

Crenella decussate/, Mont. 

Mytilus decussatus, Montagu, Test. Brit. Suppl., p. 69 (1809). 

,, ,, Forbes and Hanley, ii. p. 210, pi. xlv. fig. 2. 

Crenella „ Jeffreys, Brit. Conch., ii. p. 133, V., pi. xxviii. fig. 6. 

,, ,, Sowerby, III. Index Brit. Shells, pi. vii. fig. 17. 

Hab. — Burdwood Bank, south of the Falkland Islands, at 56 fathoms, December 1, 
1903. Station 346. 

Very minute specimens, not exceeding 2x2 mm., the interior beautifully pale- 
nacreous ; form precisely that of the European and Canadian type, the divaricating 
sculpture seemingly also identical, as well as the fine marginal crenellations. 

Modiolarca mesembrina, M. and St. 

Modiolarca mesembrina, Melvill and Standen, Trans. Roy. Soc. Edin., vol. xlvi., part i., 

p. 146, pi. figs. 15, 15a (1907). 
Modiolarca picturala, Cooper and Preston, Ann. and Mag. N. Hist., ser. viii , vol. v., pi. iv. 

fig. 5 (1910). 

Hab. — Falkland Islands. Station 118. 

We received lately from Mr A. P. Cobb examples of M. picturata, Cooper and 
Preston, and consider it the same as our mesembrina, from the same locality, described 
three years previously. In marking and coloration it is a most variable species : in 
form it is fairly constant. 

Order Eulamellibranchiata. 
Sub-order submytilacea. 
Family Carditidse. 

Carditella pallida, Sm. 

Carditella pallida, E. A. Smith, Proc. Zool. Soc. Lond., p. 43, pi. v. figs. 9, 96(1881). 

var. duodecim-costata, nov. (Plate, figs. 19, 19a). 

Hab. — Station 346, Burdwood Bank, at 56 fathoms. Many full-grown specimens, 
but few perfect. 



362 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

In all the specimens examined of our proposed variety, the ribs are but twelve in 
number; in typical C. pallida, Sm., they number fourteen to fifteen. The straight 
angular declivity on either side of the dorsal margin seems likewise more pronounced, 
the variety thereby assuming a more fiabellate or quasi-triangular appearance. The 
general characters of the shells are identical. As Mr Smith aptly remarks, the 
superficial aspect of Cardita flabellum, Reeve,* proves it to be nearly allied. This is 
a native of Valparaiso, Chili. 

Cardita congelascens, sp. n. (Plate, fig. 23). 

C. testa parva, trapezoids, solidula, umbonibus piominulis, insequilaterali, sequivalvi, postice dorsaliter 
recta, antice breviter arcuata, deinde ventralem usque ad marginem, leniter subrotundata, superficie 
radiatim costulata, costulis incrassatis, numero ad 21, pulchre et regulariter nodulosis, nodulis imbricatulis, 
albis, nitidis pagina intus alba, valva dextra, cardinalibus dentibus duobus crassis, sinistra dente crasso, 
elongato, praeditis 

Alt. 3, diam. 4 mm. (sp. maj.). 

Hah. — Burdwood Bank, south of the Falkland Islands, at 56 fathoms. Station 346. 

Only disassociated valves occurred of a species of Cardita which seems distinct. 
We have compared it with C. modesta, velutina, antarctica, astartoides, and other 
species of the genus inhabiting these same southern waters, and find it fails exactly to 
correspond with any of them. At the same time, we doubt if any of our examples are 
adult. Still, the character of the ribs, and the ornamentation and the general contour 
of the shell, give us hope that it may be proved eventually to have been established 
on a sound basis. The specific name alludes to the icy clime where it is endemic. 



Family Astartidse. 

Astarte magellanica, Sm. 

Aatarte magellanica, E. A. Smith, Proc. Zool. Soc. Lond., p. 41, pi. v. fig. 7 (1881). 
,, ,, ,, Journ. of Conch., iii. p. 227. 

Hob. — Burdwood Bank, south of the Falkland Islands, at 56 fathoms. Station 346. 

All disassociated valves, but some in good condition, and showing the olivaceous 
epidermis. The majority possess fewer concentric ribs than the type, but we consider 
them all referable to magellanica. The allied A. longirostra, Orb., also found in this 
region, is more pronouncedly beaked, and the ribbing is far finer. The crenulation of 
the inner margin of the valves is, as pointed out by the author of the species, another 
distinctive factor in A. magellanica. 

* Reeve, Conch. Icon., i., Cardita, pi. ix. fig. 47 (1843). 



MARINE MOLLUSCA OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 363 

Family Lucinidse. 

Diplodonta lamellata, Sm. • • 

Diplodonta lamellata, E. A. Smith, Proc. Zool. Soc. Lond., p 38, pi. v. figs 1-1 c (1881). 

Hob. — Burdwood Bank, south of the Falkland Islands, at 56 fathoms. Station 346. 

A right and left valve, hardly adult, but showing the characteristic generic 
dentition, as well as the concentric lamellar ornamentation, which led to the bestowal 
of the trivial name. These lamellae appear in our small specimens more numerous, but 
we can but believe them identical, as they agree in form, and every other detail. The 
type was discovered during the survey of H.M.S. Alert in the Straits of Magellan and 
Patagonian coast. 

Cyamium denticulatum, Sm. 

Gyamium denticulatum, E. A. Smith, Nat. Antarct. Exped. : Nat. Hist., vol. ii. p. 3, pi. iii. 
figs. 4, ib (1907). 

Hob. — Burdwood Bank, 56 fathoms. Station 346. 

A curious though minute species, conspicuous for its marginal denticulation and 
radiating impressed lines, which are seen more clearly with the aid of a lens. 

Cyamium falklandicum, M. and St. 

Cyamium falklandicum, Melvill and Standen, Journ. of Conch., ix. p. 104, pi. i. fig. 12 

(1898). 
? Cyamium iridescens, Cooper and Preston, Ann. and Mag. N.H., ser. viii., vol. v. p. 112, 

pi. iv. fig. 6 (1910). 

This is a variable species, and we consider C. iridescens, Coop, and Prest., probably 
one of its extreme forms. We have examined a large number of examples, in all stages 
of growth. The specimens collected at Hearnden Water, Station 349, are as iridescent 
as those so named by Messrs Cooper and Preston. 

Family Erycinidse. 

Kellyia cycladiformis (Desh.). 

Erycina cycladiformis, Deshayes, Trait, elem., pi. xi. figs. 6-9 ; Proc. Zool. Soc. Lond., 
p. 181 (1855). 

Hob. — Burdwood Bank, at 56 fathoms. Station 346. 

We have already recorded this (loc. cit., p. 149), but it is worthy of record that 
nearly all the subsequent specimens from the same locality that have since come into 
our hands were found living inside the valves of defunct Brachiopoda, and are therefore 
in first-class condition. Saxicavse occurred with them. 



364 DR JAMES COSMO MELVILL AND MR ROBERT STANDEN ON THE 

Kellyia magellanica, Sm. 
Kellijia magellanica, E. A. Smith, Proc. Zool. Soc. Lond., p. 41, pi. v. figs. 6, 6 a, b (1881). 

Hab. — Burdwood Bank, with K. cycladiformis (Desh.). Station 346. 
Only one perfect valve, agreeing exactly with the figure 6a above quoted. 

Davisia cobbi, Coop, and Prest. 

Davisia cobbi, J. E. Cooper and H. B. Preston, Ann. and Mag. N. Hist., ser. viii., vol. v. 
pp. 113, 114, pi. iv. figs. 9, 10 (1910). 

Hab. — Burdwood Bank, Station 346, at 56 fathoms. 

A small species with peculiar hinge. It would be unfortunately impossible, from 
the very indistinct photogravure plates, to tell the generic characteristics, and we 
wish it had been possible to figure both this and the Malvinasia, described at the same 
opportunity, in a more satisfactory fashion. 



BIBLIOGRAPHY. 

[Supplemental to the first list in Trans. Roy. Soc. Edin., xlvi. pp. 154, 155.] 

1881. Smith, Edgar Albert, "Account of the Zoological Collections made during the Survey of H.M.S. 
Alert in the Straits of Magellan, and the Coasts of Patagonia : Mollusca," P.Z.S. Lond., 
pp. 22-44, pi. iii.-v. 
[Rossia patagonica, Loligo patagonica, Onychoteuthis ingens, Pleurotoma (Bela) Cunninghami, PI. 
(Mangilia) Coppingeri, Lachesis meridionalis, Euthria atrata, E. meridionalis, Nassa (Tritia) 
Coppingeri, Lamellaria patagonica, Collonia Cunninghami, Trochus (Zizyphinus) consimilis, 
Tectura (Pilidium) Coppingeri, Chiton (Isclmochiton) imitator, Diplodonta lamellata, Mactra 
(Mulinia) levicardo, Loripes pertenuis, Kellia magellanica, Astarte magellanica, Cardita 
(Actinobolus) velutina, Carditella (n. g.) pallida, spp. nov. marinse.] 

1907. Joubin, L., Expedition Antarctique Francaise (1903-5) commandee par Dr Jean Charcot: Sciences 
Naturelles, " Documents Scientifiques : Cephalopodes," Paris, 1 pi. 

1907. Melvill, James Cosmo, and Standen, Robert, "The Marine Mollusca of the Scottish National 
Antarctic Expedition," Trans. Boy. Soc. Edin., xlvi., part i., pp. 119-157, 1 pi. 
[Tugalia antarctica, Littorina (Laivilitorina) coriacea, Lacuna notorcadensis, Rissoa Edgariana, R. 
scotiana, Cerithiopsis malvinarum, Trophon minutus, Nassa (Ilyanassa) Vallentini, Chrysodomus 
(Sipho) archibenthalis, C. (Sipho) crassicostatus, Columbarium benthocallis, Dentalium eupatrides, 
Area (Bathyarca) Strebeli, Lissarca notorcadensis, Modiolarca mesembrina, Pecten pteriola, 
Amussium \8-liratum, Lima goughensis, Kellia Lamyi (nom. nov.), Cuspidaria Brucei, Scacchia 
plenilunium, Pecten multicolor, spp. n.] 

1907. Strebel, Hermann, "Beitrage zu Kenntnis der Mollusken-Faunen der Magalhaen. Provinz," Part v., 
Zool. Jahrb. Syst. Jena, pp. 79-196, Taf. viii. 
[A continuation of the enumeration of the Molluscan fauna of the Falkland Islands, with new species 
of Megatebennus, Tugalia, Patinella, and several non-marine forms.] 

1907. Eliot, Sir Charles N. E., K.C.M.G., "Nudihranchs from New Zealand and the Falkland Isles," 
Pr